Knowledge management method and apparatus, and storage medium

Abstract

Provided in the present disclosure are a knowledge management method and apparatus, and a storage medium. The knowledge management method comprises: a first node sending first request information, the first request information being used for requesting first knowledge; and the first node receiving first response information, the first response information comprising the first knowledge.

Description

Knowledge management method and device, and storage medium

[0001] The present application claims priority to the Chinese patent application No. 202411097945.X, filed on August 9, 2024, the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] The present disclosure relates to the field of communication technology, and in particular, to a knowledge management method and device, and a storage medium. BACKGROUND

[0003] When a management service producer provides a management service (for example, an artificial intelligence / machine learning management service, a control closed-loop management service, a management data analysis service) in an actual application scenario, a series of knowledge will be generated, including configuration information of a network element, a business guarantee policy, and monitored indicators and data, and the like. The knowledge can be used to guide the application of a similar management service next time, thereby improving network operation and management efficiency. SUMMARY

[0004] In a first aspect, the present disclosure provides a knowledge management method applied to a first node. The knowledge management method comprises: sending first request information, the first request information being used to request first knowledge; and receiving first response information, the first response information comprising the first knowledge.

[0005] In an implementation manner of the above first aspect, the first knowledge comprises structured knowledge and / or unstructured knowledge.

[0006] In an implementation manner of the above first aspect, corresponding to the structured knowledge, the first knowledge comprises knowledge type, knowledge range, and / or knowledge information.

[0007] In an implementation manner of the above first aspect, corresponding to the unstructured knowledge, the first knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0008] In an implementation manner of the above first aspect, the knowledge type comprises an artificial intelligence / machine learning management type and / or a control closed-loop type.

[0009] In an implementation manner of the above first aspect, corresponding to the knowledge type of the first knowledge being the artificial intelligence / machine learning management type, the knowledge information comprises knowledge unit information of at least one machine learning workflow stage. The machine learning workflow stage comprises at least one of the following: a machine learning model training stage, a machine learning model testing stage, an artificial intelligence / machine learning inference simulation stage, a machine learning model deployment stage, and an artificial intelligence / machine learning inference stage.

[0010] In an implementation form of the first aspect, corresponding to the machine learning workflow stage being a machine learning model training stage, the knowledge unit information comprises at least one of training data information, training strategy information, and machine learning model information. The training data information comprises at least one of training data feature information, training data distribution information, training data volume information, and training data collection strategy. The training strategy information comprises training energy consumption information, and / or training orchestration information. The machine learning model information comprises at least one of model energy consumption information, model complexity information, model size information, model performance information, and model confidence information.

[0011] In an implementation form of the first aspect, corresponding to the machine learning workflow stage being a machine learning model testing stage, the knowledge unit information comprises at least one of test set to training set ratio information, testing data information, and testing strategy information.

[0012] In an implementation form of the first aspect, corresponding to the machine learning workflow stage being an artificial intelligence / machine learning inference stage, the knowledge unit information comprises at least one of inference data information, inference strategy information, and machine learning model information. The inference data information comprises at least one of inference data feature information, inference data distribution information, inference data volume information, inference data collection strategy, and data collection condition. The inference strategy information comprises inference energy consumption information, and / or inference orchestration information.

[0013] In an implementation form of the first aspect, corresponding to the knowledge type of the first knowledge being a control loop type, the knowledge information comprises at least one of control loop category, control loop business category, control loop component information, and control loop corresponding knowledge unit information.

[0014] In an implementation form of the first aspect, the control loop corresponding knowledge unit information comprises knowledge unit information corresponding to at least one control loop stage. The control loop stage comprises at least one of monitoring stage, analysis stage, decision stage, and execution stage.

[0015] In an implementation form of the first aspect, corresponding to the control loop stage being a monitoring stage, the control loop stage corresponding knowledge unit information comprises at least one of monitoring indicator and monitoring strategy.

[0016] In an implementation form of the first aspect, corresponding to the control loop stage being an analysis stage or a decision stage, the control loop stage corresponding knowledge unit information comprises at least one of control loop fault information, control loop conflict information, and fault handling suggestion.

[0017] In an implementation form of the first aspect, the control loop stage corresponds to an execution stage, and the knowledge unit information corresponding to the control loop stage comprises scene information, an execution scheme, an execution board, and execution flow information.

[0018] In an implementation form of the first aspect, the first request message comprises at least one of the following: a knowledge type of the first knowledge, a knowledge range, knowledge information, and knowledge filtering information.

[0019] In an implementation form of the first aspect, the first node is a management service supply consumer node or a management service supply producer node.

[0020] In an implementation form of the first aspect, the first node is a management service supply producer node, and the knowledge management method further comprises: creating the first instance based on the first response information; and sending a creation response information to a third node. The third node is a management service supply consumer node.

[0021] In an implementation form of the first aspect, before the first request information is sent, the knowledge management method further comprises: receiving a creation request from a third node. The creation request is used for the management service supply producer node to create the first instance.

[0022] In an implementation form of the first aspect, the first node is a management service supply consumer node, and the knowledge management method further comprises: determining a creation request based on the first response information, the creation request being used for creating the first instance; sending the creation request to a fourth node, the fourth node being a management service supply producer node; and receiving a creation response information from the fourth node.

[0023] In an implementation form of the first aspect, the knowledge type of the first knowledge is an artificial intelligence / machine learning management type, and the first instance is an artificial intelligence / machine learning instance.

[0024] In an implementation form of the first aspect, the knowledge type of the first knowledge is a control loop type, and the first instance is a control loop instance.

[0025] In an implementation form of the first aspect, the knowledge management method further comprises: sending second request information, the second request information being used for requesting an archiving operation on the second knowledge of the first instance; and receiving second response information, the second response information comprising an archiving state of the second knowledge.

[0026] In an implementation form of the first aspect, the second knowledge comprises structured knowledge and / or unstructured knowledge.

[0027] In conjunction with the first aspect mentioned above, in one implementation method, corresponding to structured knowledge, the second knowledge includes at least one of the following: knowledge type, knowledge scope, and knowledge information.

[0028] In conjunction with the first aspect mentioned above, in one implementation method, corresponding to unstructured knowledge, the second knowledge is the knowledge vector and the index information corresponding to the knowledge vector.

[0029] Secondly, this disclosure provides another knowledge management method applied to a second node. This knowledge management method includes: receiving first request information, the first request information being used to request first knowledge; and sending first response information, the first response information including the first knowledge.

[0030] In conjunction with the second aspect above, in one implementation, the first knowledge includes structured knowledge and / or unstructured knowledge.

[0031] In conjunction with the second aspect above, in one implementation method, corresponding to structured knowledge, the first knowledge includes knowledge type, knowledge scope, and / or knowledge information.

[0032] In conjunction with the second aspect mentioned above, in one implementation method, corresponding to unstructured knowledge, the first knowledge is the knowledge vector and the index information corresponding to the knowledge vector.

[0033] In conjunction with the second aspect above, in one implementation, the knowledge type includes: artificial intelligence / machine learning management type, and / or, control loop type.

[0034] In conjunction with the second aspect above, in one implementation, the knowledge type corresponding to the first knowledge is the artificial intelligence / machine learning management type, and the knowledge information includes knowledge unit information of at least one machine learning workflow stage. The machine learning workflow stage includes at least one of the following: machine learning model training stage, machine learning model testing stage, artificial intelligence / machine learning inference simulation stage, machine learning model deployment stage, and artificial intelligence / machine learning inference stage.

[0035] In conjunction with the second aspect above, in one implementation, the corresponding stage of the machine learning workflow is the machine learning model training stage. The knowledge unit information includes at least one of the following: training data information, training strategy information, and machine learning model information. The training data information includes at least one of the following: training data feature information, training data distribution information, training data volume information, and training data collection strategy; the training strategy information includes training energy consumption information and / or training orchestration information; the machine learning model information includes at least one of the following: model energy consumption information, model complexity information, model size information, model performance information, and model confidence information.

[0036] In an implementation form of the second aspect, the machine learning workflow stage corresponding to the first knowledge is a machine learning model testing stage, and the knowledge unit information comprises at least one of testing set and training set ratio information, testing data information, and testing strategy information.

[0037] In an implementation form of the second aspect, the machine learning workflow stage corresponding to the first knowledge is an artificial intelligence / machine learning inference stage, and the knowledge unit information comprises at least one of inference data information, inference strategy information, and machine learning model information. The inference data information comprises at least one of inference data feature information, inference data distribution information, inference data quantity information, inference data collection strategy, and data collection condition. The inference strategy information comprises inference energy consumption information and / or inference orchestration information.

[0038] In an implementation form of the second aspect, the knowledge type of the first knowledge is a control closed loop type, and the knowledge information comprises at least one of control closed loop category, control closed loop business category, control closed loop component information, and control closed loop corresponding knowledge unit information.

[0039] In an implementation form of the second aspect, the control closed loop corresponding knowledge unit information comprises knowledge unit information corresponding to at least one control closed loop stage. The control closed loop stage comprises at least one of a monitoring stage, an analysis stage, a decision stage, and an execution stage.

[0040] In an implementation form of the second aspect, the control closed loop stage corresponding to the control closed loop corresponding knowledge unit information is a monitoring stage, and the control closed loop stage corresponding knowledge unit information comprises at least one of monitoring indicators and monitoring strategies.

[0041] In an implementation form of the second aspect, the control closed loop stage corresponding to the control closed loop corresponding knowledge unit information is an analysis stage or a decision stage, and the control closed loop stage corresponding knowledge unit information comprises at least one of control closed loop fault information, control closed loop conflict information, and fault handling suggestions.

[0042] In an implementation form of the second aspect, the control closed loop stage corresponding to the control closed loop corresponding knowledge unit information is an execution stage, and the control closed loop stage corresponding knowledge unit information comprises scene information, execution scheme, execution board, and execution flow information.

[0043] In an implementation form of the second aspect, the first request message comprises at least one of the knowledge type of the first knowledge, the knowledge range, the knowledge information, and the knowledge filtering information.

[0044] In an implementation form of the second aspect, the knowledge management method further includes: receiving second request information, the second request information being used to request an archiving operation on the second knowledge of the first instance; and sending second response information, the second response information including an archiving state of the second knowledge.

[0045] In an implementation form of the second aspect, the second knowledge includes structured knowledge and / or unstructured knowledge.

[0046] In an implementation form of the second aspect, corresponding to the structured knowledge, the second knowledge includes at least one of: a knowledge type, a knowledge range, and knowledge information.

[0047] In an implementation form of the second aspect, corresponding to the unstructured knowledge, the second knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0048] In a third aspect, the present disclosure provides another knowledge management method, applied to a third node. The knowledge management method includes: sending a creation request to a fourth node, the creation request being used to instruct the fourth node to create a control loop instance based on first knowledge. The fourth node is a management service supply producer node.

[0049] In an implementation form of the third aspect, the control loop instance includes a control loop monitoring instance and / or a control loop management instance.

[0050] In an implementation form of the third aspect, parameters of the control loop monitoring instance include at least one of: a control loop monitoring indicator, a control loop monitoring strategy, and a control loop target.

[0051] In an implementation form of the third aspect, parameters of the control loop management instance include at least one of: a control loop monitoring indicator, a control loop monitoring strategy, a control loop target, a control loop priority, a control loop category, a control loop service category, a control loop component, a control loop object, and a control loop action condition.

[0052] In an implementation form of the third aspect, the knowledge management method further includes: receiving creation response information from the fourth node, the creation response information including parameters of the control loop instance, a problem to be monitored for the control loop instance, and a processing suggestion corresponding to the problem.

[0053] In combination with the third aspect, in an implementation form, the knowledge management method further comprises: monitoring a control loop monitoring indicator according to the parameters in the control loop monitoring instance and / or the control loop management instance, and creating a control loop report. The control loop report comprises at least one of a control loop operation report, a control loop fault report, and a control loop conflict report.

[0054] In combination with the third aspect, in an implementation form, the control loop fault report comprises at least one of abnormal indicator information, fault category information, fault entity information, fault positioning information, fault cause information, and a fault recommended solution.

[0055] In combination with the third aspect, in an implementation form, the control loop conflict report comprises at least one of a conflict category, a conflict cause, and a conflict recommended solution.

[0056] In combination with the third aspect, in an implementation form, the control loop operation report comprises at least one of parameters of the control loop instance, the control loop monitoring indicator, a control loop target completion condition, and a fault handling condition.

[0057] In combination with the third aspect, in an implementation form, the knowledge management method further comprises: sending modification request information to the fourth node, the modification request being used to indicate that the control loop instance is modified according to the control loop report; and receiving modification response information from the fourth node.

[0058] In the fourth aspect, the disclosure provides another knowledge management method, applied to a fourth node. The knowledge management method comprises: creating a control loop instance based on first knowledge; and sending creation response information to a third node. The third node is a management service provider node.

[0059] In combination with the fourth aspect, in an implementation form, the knowledge management method further comprises: receiving a dynamic control loop creation request from the third node. The dynamic control loop request comprises one of the following parameters: dynamic control loop indication information, used to indicate that the control loop is a dynamic control loop; dynamic control loop component information; a dynamic control loop monitoring indicator; and a dynamic control loop target.

[0060] In combination with the fourth aspect, in an implementation form, the knowledge management method further comprises: modifying the control loop instance according to the monitoring condition of the control loop monitoring indicator; and sending a dynamic control loop report to the third node.

[0061] In a fifth aspect, the present disclosure provides a knowledge management apparatus, comprising: a communication unit and a processing unit. The communication unit is configured to send first request information, wherein the first request information is used to request first knowledge. The communication unit is further configured to receive first response information, wherein the first response information comprises the first knowledge.

[0062] In an implementation form of the above fifth aspect, the first knowledge comprises structured knowledge and / or unstructured knowledge.

[0063] In an implementation form of the above fifth aspect, corresponding to the structured knowledge, the first knowledge comprises knowledge type, knowledge scope, and / or knowledge information.

[0064] In an implementation form of the above fifth aspect, corresponding to the unstructured knowledge, the first knowledge comprises a knowledge vector and index information corresponding to the knowledge vector.

[0065] In an implementation form of the above fifth aspect, the knowledge type comprises artificial intelligence / machine learning management type and / or control closed loop type.

[0066] In an implementation form of the above fifth aspect, corresponding to the knowledge type of the first knowledge being the artificial intelligence / machine learning management type, the knowledge information comprises knowledge unit information of at least one machine learning workflow stage. The machine learning workflow stage comprises at least one of the following: machine learning model training stage, machine learning model testing stage, artificial intelligence / machine learning inference simulation stage, machine learning model deployment stage, and artificial intelligence / machine learning inference stage.

[0067] In an implementation form of the above fifth aspect, corresponding to the machine learning workflow stage being the machine learning model training stage, the knowledge unit information comprises at least one of the following: training data information, training strategy information, and machine learning model information. The training data information comprises at least one of the following: training data feature information, training data distribution information, training data amount information, and training data collection strategy. The training strategy information comprises training energy consumption information and / or training orchestration information. The machine learning model information comprises at least one of the following: model energy consumption information, model complexity information, model size information, model performance information, and model confidence information.

[0068] In an implementation form of the above fifth aspect, corresponding to the machine learning workflow stage being the machine learning model testing stage, the knowledge unit information comprises at least one of the following: test set to training set ratio information, test data information, and test strategy information.

[0069] In an implementation form of the fifth aspect as above, the knowledge unit information comprises at least one of inference data information, inference strategy information, and machine learning model information, in correspondence with the machine learning workflow stage being an artificial intelligence / machine learning inference stage. The inference data information comprises at least one of inference data feature information, inference data distribution information, inference data volume information, inference data collection strategy, and data collection condition. The inference strategy information comprises inference energy consumption information and / or inference orchestration information.

[0070] In an implementation form of the fifth aspect as above, the knowledge type of the first knowledge is a control loop type, and the knowledge information comprises at least one of a control loop category, a control loop service category, control loop component information, and control loop corresponding knowledge unit information.

[0071] In an implementation form of the fifth aspect as above, the control loop corresponding knowledge unit information comprises knowledge unit information corresponding to at least one control loop stage. The control loop stage comprises at least one of a monitoring stage, an analysis stage, a decision stage, and an execution stage.

[0072] In an implementation form of the fifth aspect as above, in correspondence with the control loop stage being a monitoring stage, the control loop stage corresponding knowledge unit information comprises at least one of a monitoring indicator and a monitoring strategy.

[0073] In an implementation form of the fifth aspect as above, in correspondence with the control loop stage being an analysis stage or a decision stage, the control loop stage corresponding knowledge unit information comprises at least one of control loop fault information, control loop conflict information, and fault handling suggestion.

[0074] In an implementation form of the fifth aspect as above, in correspondence with the control loop stage being an execution stage, the control loop stage corresponding knowledge unit information comprises scene information, execution scheme, execution board, and execution flow information.

[0075] In an implementation form of the fifth aspect as above, the first request message comprises at least one of the knowledge type of the first knowledge, the knowledge range, the knowledge information, and the knowledge filtering information.

[0076] In an implementation form of the fifth aspect as above, the first node is a management service supply consumer node or a management service supply producer node.

[0077] In an implementation form of the fifth aspect as above, in correspondence with the first node being a management service supply producer node, the knowledge management method further comprises: creating the first instance based on the first response information; and sending a creation response information to a third node. The third node is a management service supply consumer node.

[0078] In an implementation form of the fifth aspect as above, before sending the first request information, the communication unit is further configured to receive a creation request from a third node; the creation request is used for the management service provision producer node to create the first instance.

[0079] In an implementation form of the fifth aspect as above, the processing unit is configured to determine a creation request based on the first response information; the creation request is used for creating the first instance, in response to the first node being a management service provision consumer node. The communication unit is further configured to send the creation request to a fourth node, the fourth node being a management service provision producer node; and receive a creation response information from the fourth node.

[0080] In an implementation form of the fifth aspect as above, the knowledge type corresponding to the first knowledge is an artificial intelligence / machine learning management type, and the first instance is an artificial intelligence / machine learning instance.

[0081] In an implementation form of the fifth aspect as above, the knowledge type corresponding to the first knowledge is a control closed loop type, and the first instance is a control closed loop instance.

[0082] In an implementation form of the fifth aspect as above, the communication unit is further configured to send second request information, the second request information being used for requesting an archiving operation to be performed on the second knowledge of the first instance; and receive second response information, the second response information including an archiving state of the second knowledge.

[0083] In an implementation form of the fifth aspect as above, the second knowledge includes structured knowledge, and / or unstructured knowledge.

[0084] In an implementation form of the fifth aspect as above, the second knowledge corresponding to the structured knowledge includes at least one of the following: a knowledge type, a knowledge scope, and knowledge information.

[0085] In an implementation form of the fifth aspect as above, the second knowledge corresponding to the unstructured knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0086] In a sixth aspect, the present disclosure provides another knowledge management apparatus, which comprises a communication unit. The communication unit is configured to receive first request information; the first request information being used for requesting first knowledge. The communication unit is further configured to send first response information; the first response information including the first knowledge.

[0087] In an implementation form of the sixth aspect as above, the first knowledge includes structured knowledge, and / or unstructured knowledge.

[0088] In an implementation form of the sixth aspect above, the first knowledge comprises a knowledge type, a knowledge scope, and / or knowledge information corresponding to the structured knowledge.

[0089] In an implementation form of the sixth aspect above, the first knowledge is a knowledge vector and index information corresponding to the knowledge vector corresponding to the unstructured knowledge.

[0090] In an implementation form of the sixth aspect above, the knowledge type comprises an artificial intelligence / machine learning management type, and / or a control closed loop type.

[0091] In an implementation form of the sixth aspect above, the knowledge type corresponding to the first knowledge is the artificial intelligence / machine learning management type, and the knowledge information comprises knowledge unit information of at least one machine learning workflow stage. The machine learning workflow stage comprises at least one of a machine learning model training stage, a machine learning model testing stage, an artificial intelligence / machine learning inference simulation stage, a machine learning model deployment stage, and an artificial intelligence / machine learning inference stage.

[0092] In an implementation form of the sixth aspect above, the machine learning workflow stage is the machine learning model training stage, and the knowledge unit information comprises at least one of training data information, training strategy information, and machine learning model information. The training data information comprises at least one of training data feature information, training data distribution information, training data volume information, and training data collection strategy. The training strategy information comprises training energy consumption information, and / or training orchestration information. The machine learning model information comprises at least one of model energy consumption information, model complexity information, model size information, model performance information, and model confidence information.

[0093] In an implementation form of the sixth aspect above, the machine learning workflow stage is the machine learning model testing stage, and the knowledge unit information comprises at least one of a ratio of a test set to a training set, test data information, and test strategy information.

[0094] In an implementation form of the sixth aspect above, the machine learning workflow stage is the artificial intelligence / machine learning inference stage, and the knowledge unit information comprises at least one of inference data information, inference strategy information, and machine learning model information. The inference data information comprises at least one of inference data feature information, inference data distribution information, inference data volume information, inference data collection strategy, and data collection condition. The inference strategy information comprises inference energy consumption information, and / or inference orchestration information.

[0095] In an implementation form of the sixth aspect above, the knowledge type corresponding to the first knowledge is a control loop type, and the knowledge information includes at least one of a control loop category, a control loop service category, control loop component information, and control loop corresponding knowledge unit information.

[0096] In an implementation form of the sixth aspect above, the control loop corresponding knowledge unit information includes at least one control loop phase corresponding knowledge unit information. The control loop phase includes at least one of a monitoring phase, an analysis phase, a decision phase, and an execution phase.

[0097] In an implementation form of the sixth aspect above, the control loop phase corresponding knowledge unit information corresponding to the monitoring phase includes at least one of a monitoring index and a monitoring strategy.

[0098] In an implementation form of the sixth aspect above, the control loop phase corresponding knowledge unit information corresponding to the analysis phase or the decision phase includes at least one of control loop fault information, control loop conflict information, and fault handling suggestions.

[0099] In an implementation form of the sixth aspect above, the control loop phase corresponding knowledge unit information corresponding to the execution phase includes scene information, an execution scheme, an execution board, and execution flow information.

[0100] In an implementation form of the sixth aspect above, the first request message includes at least one of a knowledge type of the first knowledge, a knowledge range, knowledge information, and knowledge filtering information.

[0101] In an implementation form of the sixth aspect above, the communication unit is further configured to receive second request information, the second request information being used to request an archive operation on a second knowledge of the first instance; and send second response information, the second response information including an archive state of the second knowledge.

[0102] In an implementation form of the sixth aspect above, the second knowledge includes structured knowledge and / or unstructured knowledge.

[0103] In an implementation form of the sixth aspect above, the second knowledge corresponding to the structured knowledge includes at least one of a knowledge type, a knowledge range, and knowledge information.

[0104] In an implementation form of the sixth aspect above, the second knowledge corresponding to the unstructured knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0105] In a seventh aspect, the present disclosure provides yet another knowledge management apparatus, comprising: a communication unit and a processing unit. The communication unit is configured to send a creation request to a fourth node, the creation request being configured to instruct the fourth node to create a control loop instance based on first knowledge. The fourth node is a management service provider node.

[0106] In an implementation form of the above seventh aspect, the control loop instance comprises a control loop monitoring instance and / or a control loop management instance.

[0107] In an implementation form of the above seventh aspect, the parameters of the control loop monitoring instance comprise at least one of a control loop monitoring indicator, a control loop monitoring strategy, and a control loop target.

[0108] In an implementation form of the above seventh aspect, the parameters of the control loop management instance comprise at least one of a control loop monitoring indicator, a control loop monitoring strategy, a control loop target, a control loop priority, a control loop category, a control loop service category, a control loop component, a control loop object, and a control loop action condition.

[0109] In an implementation form of the above seventh aspect, the communication unit is further configured to receive a creation response information from the fourth node, the creation response information comprising the parameters of the control loop instance, a problem to be monitored for the control loop instance, and a processing suggestion corresponding to the problem.

[0110] In an implementation form of the above seventh aspect, the processing unit is configured to monitor a control loop monitoring indicator according to the parameters in the control loop monitoring instance and / or the control loop management instance, and create a control loop report. The control loop report comprises at least one of a control loop operation report, a control loop fault report, and a control loop conflict report.

[0111] In an implementation form of the above seventh aspect, the control loop fault report comprises at least one of an abnormal indicator information, a fault category information, a fault entity information, a fault positioning information, a fault reason information, and a fault recommended solution.

[0112] In an implementation form of the above seventh aspect, the control loop conflict report comprises at least one of a conflict category, a conflict reason, and a conflict recommended solution.

[0113] In an implementation form of the above seventh aspect, the control loop operation report comprises at least one of the parameters of the control loop instance, the control loop monitoring indicator, a control loop target completion situation, and a fault processing situation.

[0114] In combination with the seventh aspect above, in an implementation form, the communication unit is further configured to send a modification request information to the fourth node, the modification request being configured to indicate that the control loop instance is modified according to the control loop report; and receive a modification response information from the fourth node.

[0115] In combination with the eighth aspect above, in an implementation form, the communication unit is further configured to receive a dynamic control loop creation request from the third node. The dynamic control loop request comprises one of the following parameters: dynamic control loop indication information configured to indicate that the control loop is a dynamic control loop; dynamic control loop component information; dynamic control loop monitoring index; and dynamic control loop target.

[0116] In combination with the eighth aspect above, in an implementation form, the processing unit is further configured to modify the control loop instance according to a monitoring result of the control loop monitoring index. The communication unit is further configured to send a dynamic control loop report to the third node.

[0117] In combination with the eighth aspect above, in an implementation form, the processing unit is further configured to modify the control loop instance according to a monitoring result of the control loop monitoring index. The communication unit is further configured to send a dynamic control loop report to the third node.

[0118] In combination with the eighth aspect above, in an implementation form, the processing unit is further configured to modify the control loop instance according to a monitoring result of the control loop monitoring index. The communication unit is further configured to send a dynamic control loop report to the third node.

[0119] In combination with the eighth aspect above, in an implementation form, the processing unit is further configured to modify the control loop instance according to a monitoring result of the control loop monitoring index. The communication unit is further configured to send a dynamic control loop report to the third node.

[0120] In combination with the eighth aspect above, in an implementation form, the processing unit is further configured to modify the control loop instance according to a monitoring result of the control loop monitoring index. The communication unit is further configured to send a dynamic control loop report to the third node.

[0121] In a twelfth aspect, the present disclosure provides a chip, which includes a processor and a communication interface. The communication interface is coupled with the processor. The processor is configured to run computer programs or instructions to implement the knowledge management method described in the first aspect, any implementation manner of the first aspect, the second aspect, any implementation manner of the second aspect, the third aspect, any implementation manner of the third aspect, the fourth aspect, or any implementation manner of the fourth aspect. For example, the chip provided in the present disclosure further includes a memory configured to store the computer programs or instructions.

[0122] It should be noted that the computer instructions described above can be stored on a computer readable storage medium in whole or in part. The computer readable storage medium can be packaged together with the processor of the device or packaged separately from the processor of the device, and the present disclosure does not limit the computer readable storage medium.

[0123] The description of the second aspect to the twelfth aspect in the present disclosure can refer to the detailed description of the first aspect.

[0124] In the present disclosure, the name of the knowledge management device described above does not constitute a limitation on the device or the functional module itself. In actual implementation, these devices or functional modules can appear in other names as long as the functions of the respective devices or functional modules are similar to those of the present disclosure, which falls within the scope of the claims of the present disclosure and equivalent technologies.

[0125] These aspects or other aspects of the present disclosure will be more apparent in the following description. BRIEF DESCRIPTION OF DRAWINGS

[0126] FIG. 1 is a schematic diagram of a workflow of artificial intelligence / machine learning according to an embodiment of the present disclosure.

[0127] FIG. 2 is a schematic diagram of an overall solution and information flow between service management and closed-loop steps according to an embodiment of the present disclosure.

[0128] FIG. 3 is a schematic diagram of a service management architecture according to an embodiment of the present disclosure.

[0129] FIG. 4 is a schematic diagram of a hardware structure of an electronic device according to an embodiment of the present disclosure.

[0130] FIG. 5 is a flowchart of a knowledge management method according to an embodiment of the present disclosure.

[0131] FIG. 6 is a flowchart of another knowledge management method according to an embodiment of the present disclosure.

[0132] FIG. 7 is a flowchart of yet another knowledge management method according to an embodiment of the present disclosure.

[0133] FIG. 8 is a flowchart of yet another knowledge management method according to an embodiment of the present disclosure.

[0134] FIG. 9 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0135] FIG. 10 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0136] FIG. 11 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0137] FIG. 12 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0138] FIG. 13 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0139] FIG. 14 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0140] FIG. 15 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0141] FIG. 16 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0142] FIG. 17 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0143] FIG. 18 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0144] FIG. 19 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0145] FIG. 20 is a flowchart of still another knowledge management method according to an embodiment of the present disclosure.

[0146] FIG. 21 is a structural schematic diagram of a knowledge management apparatus according to an embodiment of the present disclosure.

[0147] FIG. 22 is a structural schematic diagram of another knowledge management apparatus according to an embodiment of the present disclosure.

[0148] FIG. 23 is a structural schematic diagram of still another knowledge management apparatus according to an embodiment of the present disclosure.

[0149] FIG. 24 is a structural schematic diagram of still another knowledge management apparatus according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0150] With reference to the drawings, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present disclosure.

[0151] The term "and / or" in the present disclosure is merely used to describe an association relationship of associated objects, and indicates that three relationships can exist, for example, A and / or B can represent: only A, only B, and A and B.

[0152] The terms "first" and "second" and the like in the description of the present disclosure and the drawings are used to distinguish different objects or different treatments of the same object, rather than to describe a specific order of the objects.

[0153] In addition, the terms "include" and "have" and any variations thereof mentioned in the description of the present disclosure are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but can optionally include other steps or units not listed, or can optionally include other steps or units inherent to the process, method, product or device.

[0154] It should be noted that in the embodiments of the present disclosure, the words "exemplary" or "for example" are used to describe examples, illustrations, or descriptions. Any embodiment or design scheme described in the embodiments of the present disclosure through "exemplary" or "for example" should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the use of "exemplary" or "for example" is intended to present the relevant concept in a specific manner.

[0155] In the description of the present disclosure, the meaning of "a plurality of" is two or more, unless otherwise specified.

[0156] In the following, the nouns related to the embodiments of the present disclosure are explained to facilitate the reader's understanding.

[0157] (1) Artificial intelligence (artificial intelligence, AI) / machine learning (machine learning, ML)

[0158] As shown in FIG. 1, the workflow of artificial intelligence / machine learning involves four main stages, namely the training stage, the simulation stage, the deployment stage and the inference stage.

[0159] In the following, the training stage, the simulation stage, the deployment stage and the inference stage are introduced respectively.

[0160] 1. Training phase

[0161] As shown in FIG. 1, the training phase includes machine learning training and machine learning testing.

[0162] Machine learning training: training of one or a group of machine learning models, including initial training, retraining, and validation of machine learning entities to evaluate the performance of the machine learning entities when executed on training data and validation data. If the validation results are not as expected (e.g., variance is not acceptable), the machine learning model associated with the entity needs to be retrained. Machine learning model training is the initial stage of the workflow.

[0163] Machine learning testing: testing of validated machine learning entities to evaluate the performance of the trained machine learning models when executed on test data. If the test results meet expectations, the machine learning entity can proceed to the next phase, otherwise the machine learning model associated with the entity may need to be retrained.

[0164] 2. Simulation phase

[0165] Running machine learning entities in a simulation environment for machine learning simulation, inference. The purpose is to evaluate the inference performance of machine learning entities in the simulation environment before applying machine learning entities to target networks or systems.

[0166] In the workflow of artificial intelligence / machine learning, the simulation phase is optional and can be skipped in the workflow of artificial intelligence / machine learning.

[0167] 3. Deployment phase

[0168] The deployment phase mainly includes machine learning entity loading, i.e., the process of making the trained machine learning entity available for target artificial intelligence / machine learning inference functions (also known as atomic operation series). In some cases, the deployment phase may not be needed, for example, when the training function and the inference function are located in the same place, then the deployment phase is not needed.

[0169] 4. Inference phase

[0170] Through artificial intelligence / machine learning inference, the trained machine learning entity performs inference.

[0171] (2) Control closed control loop (CCL)

[0172] Communication service assurance relies on a set of management services that provide a communications service provider (CSP) with the ability to ensure the communication service according to the agreement (e.g., service level specification (SLS) agreement) signed with the communications service consumer (CSC) (e.g., enterprise).

[0173] Figure 2 shows the overall solution and information flow between management services and closed-loop steps.

[0174] A controlled entity represents the resources used by the communication service, the assurance of which is provided by the control closed loop between different management services provided by the management system.

[0175] The control closed loop includes four steps: monitoring, analysis, decision, and execution. The input of the control closed loop is data about the resources used by the communication service and the corresponding service (e.g., performance measurements, key performance indicators (KPIs)). Data collection is monitored by the monitoring step of the control closed loop; the analysis service is implemented by the analysis step of the control closed loop; the decision is made by the decision support service of the control closed loop, which can be an execution action of the execution step of the control closed loop, and the decision step provides different levels of automated decision-making and manual supervision support; the orchestration or control step of the control closed loop implements the provisioning service.

[0176] The monitoring step sends the monitored information to the analysis step, which outputs a database, which in turn is used by the decision step to make decisions, which are then controlled by the execution step to control the controlled entity to execute the action, and output data to the monitoring step.

[0177] (3) Retrieval-augmented generation (RAG) technology

[0178] Retrieval-augmented generation (RAG) technology is a method in the field of natural language processing technology that combines the advantages of retrieval technology and generation technology.

[0179] For example, when generating text, a retrieval-augmented generation model not only relies on the internal knowledge of a pre-trained language model, but also retrieves relevant documents from an external database to enhance the accuracy and diversity of the generated results.

[0180] Below, the workflow, advantages, and application scenarios of the retrieval-augmented generation model are introduced.

[0181] 1. Workflow of retrieval-augmented generation model

[0182] The workflow of the retrieval-augmented generation model mainly includes the retrieval phase and the generation phase.

[0183] Retrieval phase: The retrieval-augmented generation model retrieves the most relevant document fragments from the external database based on the given input query. This step usually uses information retrieval techniques, such as the best matching (BM25) retrieval technique or vector retrieval technique.

[0184] Generation phase: The model inputs the input and the retrieved document fragments into the generation model. The generation model combines the input and the retrieved document fragments to generate more rich and accurate output.

[0185] 2. Advantages of retrieval-augmented generation model

[0186] The retrieval-augmented generation model has the advantages of rich knowledge, accurate information, and high flexibility.

[0187] Rich knowledge: By retrieving external databases, the model's knowledge base can be greatly expanded, avoiding the limitations of relying solely on pre-training data.

[0188] Accurate information: The relevant document fragments retrieved can provide information support, improving the accuracy of generated text.

[0189] High flexibility: Different types of external information can be dynamically retrieved according to different tasks and requirements, making it highly adaptable.

[0190] 3. Application scenarios of retrieval-augmented generation model

[0191] The retrieval-augmented generation model can be applied to question answering scenarios, content generation scenarios, and dialogue scenarios.

[0192] Question answering scenario: By retrieving relevant documents, more accurate and detailed answers can be provided.

[0193] Content generation scenario: When generating articles, reports, and other content, relevant materials can be cited to improve content quality.

[0194] Dialogue scenario: In the dialogue process, external information can be introduced to enrich the dialogue content.

[0195] (4) Service management architecture (service based management architecture)

[0196] As shown in FIG. 3, the service management architecture includes a business support system (BSS), a cross domain management function (CD-MnF) network element, a domain management function (domain-MnF) unit, and a network element (NE).

[0197] The cross domain management function network element is used to manage one or more domain management function units. The domain management function unit can be used to manage one or more network elements.

[0198] Hereinafter, the business support system, the cross domain management function network element, the domain management function unit, and the network element are introduced respectively.

[0199] 1. Business support system

[0200] The business support system is a system for communication services, which is used to provide functions and management services such as charging, settlement, accounting, customer service, business, network monitoring, communication service lifecycle management, service intent translation, etc. The business support system can be an operational system of an operator, or a vertical operational technology system.

[0201] 2. Cross domain management function network element

[0202] The cross domain management function network element is also called a network management function (NMF) unit, which can be a network management system (NMS), a network management service producer (MnS producer), a network management service consumer (MnS consumer), a network function management service consumer (NFMS_C), etc.

[0203] The cross-domain management function network element provides one or more of the following management functions or management services: network lifecycle management, network deployment, network fault management, network performance management, network configuration management, network assurance, network optimization function, and translation of intent from a communication service provider (intent-CSP), etc. The network referred to in the above management functions or management services can include one or more network elements or sub-networks, or a network slice. That is, the network management function unit can be a network slice management function (NSMF) unit, or a cross-domain management data analysis function (MDAF) unit, or a cross-domain self-organization network function (SON function) unit, or a cross-domain intent-driven management function (intent driven MnS) unit.

[0204] 3. Domain management function unit

[0205] The domain management function unit, also referred to as a network subnet management function (NSMF) unit or a network element management function unit, can be a wireless automation engine (mobile broadband (MBB) automation engine (MAE)), an element management system (EMS), a network function management service provider (NFMS P), a network slice subnet management function (NSSMF) unit, a domain management data analysis function (domain MDAF) unit, a domain self-organization network function (SON function), a domain intent-driven management function unit, a network management service producer, a network management service consumer, and the like.

[0206] The domain management function unit can be classified in the following manner 1 and manner 2.

[0207] Manner 1

[0208] According to network type, the domain management function unit can be classified into: a radio access network (RAN) domain management function (RAN domain MnF) unit, a core network domain management function (CN domain MnF) unit, a transport network domain management function (TN domain MnF) unit, and the like.

[0209] It should be noted that the domain management function unit can also be a certain domain network management system, which can manage one or more of the access network, the core network, or the transport network.

[0210] Manner 2

[0211] According to administrative region, the domain management function unit can be classified into: a domain management function unit of a certain region, such as an A city domain management function unit, a B city domain management function unit, and the like.

[0212] The domain management function unit provides one or more of the following functions or management services: lifecycle management of a subnetwork or a network element, deployment of a subnetwork or a network element, fault management of a subnetwork or a network element, performance management of a subnetwork or a network element, assurance of a subnetwork or a network element, optimization function of a subnetwork or a network element, and translation of an intent from a network operator (Intent-NOP) of a subnetwork or a network element, and the like. The subnetwork herein includes one or more network elements. The subnetwork can also include a subnetwork, i.e., one or more subnetworks form a larger subnetwork. The subnetwork herein can also be a network slice subnetwork.

[0213] 4. Network element

[0214] The network element is an entity for providing network services. The network element includes a core network element, a radio access network element, or a transport network element, and the like.

[0215] For example, the core network network element can include, but is not limited to, an access and mobility management function (AMF) entity, a session management function (SMF) entity, a policy control function (PCF) entity, a network data analysis function (NWDAF) entity, a network repository function (NRF) entity, a gateway, etc.

[0216] The radio access network network element can include, but is not limited to, various types of base stations (for example, a next-generation node B (gNB), an evolved node B (eNB), a central unit control panel (CUCP), a central unit (CU), a distributed unit (DU), a central unit user panel (CUUP), etc.

[0217] In the present disclosure, a network function (NF) is also referred to as a network element (NE). The network element can provide one or more of the following management functions or management services: lifecycle management of the network element, deployment of the network element, fault management of the network element, performance management of the network element, assurance of the network element, optimization function of the network element, and translation of the intent of the network element, etc.

[0218] The above-mentioned management services mainly include four types of management services: a provisioning management service (prov MnS), a fault supervision management service (fault supervision MnS), a streaming data reporting management service (streaming data reporting MnS), and a file data reporting management service (file data reporting MnS).

[0219] The management service capability providing management services can include at least one of management data control, fault management services, file management services, new radio configuration, 5G core network configuration, network slice configuration, edge computing configuration, artificial intelligence / machine learning management services, management data analysis (MDA) services, self-organized network (SON) policies, access network self-configuration management services, intent-driven management services, registration and discovery management services, communication service assurance, energy entity network and service operation, access control for management service (MSAC), and control loop management services.

[0220] The cross-domain management function network element can support model training of an artificial intelligence / machine learning model, model inference of the artificial intelligence / machine learning model, and creation of a control loop instance. The domain management function unit can support model training of an artificial intelligence / machine learning model, model inference of the artificial intelligence / machine learning model, and creation of a control loop, and can also serve as an entity of the control loop. The network element, which is an entity for providing network services, can support model training of an artificial intelligence / machine learning model, model inference of the artificial intelligence / machine learning model, and can also serve as an entity of the control loop.

[0221] The above, the terms related to the embodiments of the present disclosure are explained in detail.

[0222] When the management service producer provides management services (for example, artificial intelligence / machine learning management services, control loop management services, and management data analysis services) in actual application scenarios, a series of knowledge can be generated, including configuration information of network elements, service assurance policies, and monitored indicators and data, and the like. These pieces of knowledge can be used to guide learning for the next similar management service application, and improve network operation and management efficiency.

[0223] For example, taking the artificial intelligence / machine learning management service as an example, when the machine learning model training is performed for the management data analysis type (MDA type), the knowledge involved in the machine learning model training life cycle, such as data collection, training set and test set division, training resource allocation, training node configuration, and model performance, can be used to guide the next model training, to support more efficient model training and improve the performance of the model.

[0224] Taking a control closed-loop management service as an example, in a case that the control closed-loop management service is used for business assurance, the knowledge of entities selection, index monitoring, and index threshold setting in the closed loop involved in each link of the control closed-loop management service can be used as the knowledge of the control closed-loop management service. The knowledge of the control closed-loop management service can be used to guide the control closed-loop management service to perform the next business assurance, so as to support a higher assurance performance.

[0225] However, there is no management scheme for the knowledge related to the management service and no scheme for utilizing the knowledge related to the management service in the current third generation partnership project, thereby causing problems of low efficiency and resource waste.

[0226] In view of this, the present disclosure provides a knowledge management method. A first node sends first request information, and the first request information is used to request first knowledge. The first node receives first response information, and the first response information includes the first knowledge. Compared with the current scheme for not utilizing the knowledge related to the management service, thereby causing the problem of resource waste, the first node sends the request message and receives the first response information including the first knowledge in the above technical scheme, so that the knowledge can be effectively utilized and resource waste can be avoided.

[0227] The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0228] In an implementation manner, the first node can be a management service supply consumer node or a management service supply producer node, or can be a node with other identities. The second node can be a knowledge management service producer node. The third node is a management service supply producer node. The fourth node is a management service supply consumer node.

[0229] In some embodiments, the second node can be a server or a knowledge base.

[0230] In some embodiments, the first node, the second node, the third node, and the fourth node can be a cross-domain management function network element, a domain management function unit, or a network element, and the present disclosure does not make any limitation on this.

[0231] For the cross-domain management function network element, the domain management function unit, and the network element, refer to the part of (4) service based management architecture in the above explanation of the terms related to the embodiments of the present disclosure, and details are not repeated here.

[0232] When implemented by hardware, each module in the communication system can be integrated and implemented on the hardware structure of the electronic device as shown in FIG. 4. For example, as shown in FIG. 4, the basic hardware structure of the electronic device is introduced.

[0233] FIG. 4 is a schematic diagram of a hardware structure of an electronic device according to an embodiment of the present disclosure. As shown in FIG. 4, the electronic device includes at least one processor 401, a communication line 402, and at least one communication interface 404, and can further include a memory 403. The processor 401, the memory 403, and the communication interface 404 can be connected through the communication line 402.

[0234] The processor 401 can be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement one or more embodiments of the present disclosure, for example, one or more digital signal processors (DSPs), or, one or more field programmable gate arrays (FPGAs).

[0235] The communication line 402 can include a path for transmitting information between the above-mentioned components.

[0236] The communication interface 404, which is configured to communicate with other devices or communication networks, can use any transceiver type device, for example, Ethernet, a radio access network (RAN), a wireless local area network (WLAN), etc.

[0237] The memory 403 can be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, a magnetic disk storage or other magnetic storage devices, or any other medium capable of storing desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0238] In one design, the memory 403 can be external to the processor 401, i.e., the memory 403 can be an external memory to the processor 401. In this case, the memory 403 can be connected to the processor 401 via the communication line 402, and be configured to store the execution instructions or application codes, and be controlled by the processor 401 to perform the knowledge management method provided by the embodiments of the present disclosure. In another design, the memory 403 can be integrated with the processor 401, i.e., the memory 403 can be an internal memory to the processor 401. For example, the memory 403 can be a cache memory, and be configured to temporarily store some data and instruction information, etc.

[0239] As an implementation, the processor 401 can include one or more CPUs, e.g., the CPU 0 and the CPU 1 in FIG. 4. As another implementation, the electronic device can include multiple processors, e.g., the processor 401 and the processor 407 in FIG. 4. As yet another implementation, the electronic device can further include the output device 405 and the input device 406.

[0240] It should be noted that the embodiments of the present disclosure can be mutually referenced or used as reference to each other, e.g., the same or similar steps, method embodiments, system embodiments and device embodiments can be mutually referenced, without limitation.

[0241] As an embodiment of the present disclosure, FIG. 5 is a flowchart of a knowledge management method according to an embodiment of the present disclosure, which can be applied to a first node. As shown in FIG. 5, the knowledge management method includes the following S501 and S502.

[0242] In S501, the first node sends first request information.

[0243] The first request information is used to request first knowledge.

[0244] In some embodiments, the first node can be a management service supply consumer node or a management service supply producer node, or can be a node with other identities.

[0245] In the following, the first knowledge is described.

[0246] In some embodiments, the first knowledge includes structured knowledge, and / or unstructured knowledge.

[0247] Corresponding to the unstructured knowledge, the first knowledge is a knowledge vector and index information corresponding to the knowledge vector (e.g., embedding model index information, used to generate the knowledge vector or restore the knowledge vector to knowledge. The knowledge can be structured knowledge, or knowledge represented in natural language).

[0248] It should be noted that the knowledge can be a sentence described in natural language (for example, the data distribution of the model is a target type distribution), which can be converted into a knowledge vector through an embedding model. Therefore, through the index information, the embedding model can be determined, and then through the embedding model, the knowledge vector can be converted back to the sentence described in natural language. Similarly, the knowledge vector can also be converted back to structured knowledge.

[0249] Corresponding to the structured knowledge, the first knowledge includes a knowledge type, a knowledge range, and / or knowledge information.

[0250] In some embodiments, the knowledge type includes at least one of an artificial intelligence / machine learning management type and a control closed loop type. Of course, the above is only an exemplary description of the knowledge type, and the knowledge type can also include at least one of a data control type, a fault type, a file type, a new radio configuration type, a 5G core network configuration type, a network slice configuration type, an edge computing configuration type, a data analysis type, a self-organizing network policy type, an access network self-configuration type, an intent-driven type, a registration and discovery type, a communication service assurance type, an energy consumption entity network and service operation type, without limitation of the present disclosure.

[0251] In some embodiments, the knowledge range includes at least one of access network device knowledge, core network knowledge, cross-domain knowledge, coverage type knowledge, energy consumption type knowledge, service assurance knowledge, and model reasoning type (management data analysis type, analysis identification).

[0252] In some embodiments, the knowledge information includes at least one of historical configuration information (such as configuration objects, configuration parameters, configuration values, and change situations) and historical performance information (such as performance measurement objects, collected PM / KPIs, values, and change situations).

[0253] In some embodiments, the knowledge type corresponding to the first knowledge is an artificial intelligence / machine learning management type, and the knowledge information includes knowledge unit information of at least one machine learning workflow stage.

[0254] The machine learning workflow stage includes at least one of a machine learning model training stage, a machine learning model testing stage, an artificial intelligence / machine learning inference simulation stage, a machine learning model deployment stage, and an artificial intelligence / machine learning inference stage.

[0255] Hereinafter, knowledge unit information corresponding to different machine learning workflow stages is introduced through Examples 11 to 13. Example 11 shows knowledge unit information corresponding to a machine learning model training stage; Example 12 shows knowledge unit information corresponding to a machine learning model testing stage; and Example 13 shows knowledge unit information corresponding to an artificial intelligence / machine learning inference stage.

[0256] In Example 11, corresponding to the machine learning workflow stage being a machine learning model training stage, the knowledge unit information includes at least one of: training data information, training strategy information, and machine learning model information.

[0257] The training data information includes at least one of: training data characteristic information (e.g., performance measurement, key performance indicator, key quality indicator, key experience indicator, etc.), training data distribution information (e.g., normal distribution, uniform distribution), training data volume information (e.g., training data volume information per characteristic, data volume information per training task, training data volume per node), training data collection strategy (including data collection source (e.g., network element, access network device such as base station, external data, etc.) and data collection condition (e.g., data collection condition based on a specific indicator (e.g., data collection when energy consumption is less than a specific threshold), data collection frequency, time interval, and time window)).

[0258] The training strategy information includes training energy consumption information, and / or training orchestration information (e.g., training node information, computing power information of each training node, training time information, model joint training information (e.g., model identification of joint training)).

[0259] The machine learning model information includes at least one of: model context information, model energy consumption information, model complexity information (e.g., number of floating point operations per second), model size information, model performance information, model confidence information, input distribution and output distribution of executing model inference, and association between input and output.

[0260] The model complexity information includes at least one of: number of parameters, model size, training time, inference time, million poincare-zeta function number, depth of decision tree, depth and width of neural network, capacity of model, regularization coefficient, cross-validation score.

[0261] In Example 12, corresponding to the machine learning workflow stage being a machine learning model testing stage, the knowledge unit information includes at least one of: test set to training set ratio information, test data information, and test strategy information.

[0262] In Example 13, corresponding to the machine learning workflow stage being an artificial intelligence / machine learning inference stage, the knowledge unit information includes at least one of: inference data information, inference strategy information, and machine learning model information.

[0263] The inference data information includes at least one of inference data feature information, inference data distribution information, inference data quantity information, inference data collection strategy, and data collection condition. The inference data herein includes both input data of inference and output result of inference. The inference strategy information includes inference energy consumption information and / or inference orchestration information. The inference data information is similar to the training data information, except that the application stage is different. The inference orchestration information includes model scheme used by the inference scheme, including used model and joint inference information of the model (inference order of the model, model identifier).

[0264] The above introduces the knowledge unit information corresponding to different machine learning workflow stages.

[0265] In some embodiments, the knowledge type corresponding to the first knowledge is a control closed loop type, and the knowledge information includes at least one of a control closed loop category (for example, business assurance), a control closed loop business category (for example, data-game), control closed loop component information (for example, multi-system network facility, management service, network function, cell, access network device), and control closed loop corresponding knowledge unit information.

[0266] It can be understood that the control closed loop corresponding knowledge unit information can exist as a whole, or can exist separately from a certain control closed loop stage.

[0267] The control closed loop corresponding knowledge unit information includes at least one control closed loop stage corresponding knowledge unit information. The control closed loop stage includes at least one of the following: a monitoring stage, an analysis stage, a decision stage, and an execution stage.

[0268] The following introduces the knowledge unit information corresponding to different control closed loop stages through examples 21 to 23. Example 21 shows the knowledge unit information corresponding to the monitoring stage; example 22 shows the knowledge unit information corresponding to the analysis stage or the decision stage; and example 23 shows the knowledge unit information corresponding to the execution stage.

[0269] In example 21, the control closed loop stage corresponding to the knowledge unit information is the monitoring stage, and the control closed loop stage corresponding knowledge unit information includes at least one of monitoring indicators (for example, performance measurement, key performance indicator, key quality indicator, key experience indicator, and the like feature information and its numerical range) and monitoring strategy (for example, frequency of indicator reporting, monitoring time length, and time window).

[0270] In example 22, the control closed loop stage corresponding to the knowledge unit information is the analysis stage or the decision stage, and the control closed loop stage corresponding knowledge unit information includes at least one of control closed loop fault information, control closed loop conflict information, and fault handling suggestion.

[0271] The control loop fault information includes at least one of a fault type (for example, coverage, uplink and downlink rate, anomaly index, throughput, energy consumption, service level agreement guarantee), control loop component fault performance measurement fault, and key performance indicator fault. The control loop conflict information includes at least one of a conflict type and a conflict report. The fault processing suggestion includes a service or network indicator degradation problem handling suggestion (including: problem description, problem cause analysis, problem processing suggestion). The problem processing suggestion is a configuration for the managed object, including managed object information, a configuration category to be issued, a configuration parameter, and a configuration condition.

[0272] In Example 23, the control loop phase corresponding to the execution phase, the knowledge unit information corresponding to the control loop phase includes: scene information, execution scheme, execution board, and execution flow information.

[0273] It can be understood that the guarantee execution scheme is different for different scenes (for example, control loop types and control loop service types). For example, the template of the execution scheme of a concert scene is different from the template of the execution scheme of a marathon scene, and knowledge can be recycled according to guarantee execution experience. The execution scheme is a configuration for the managed object, including managed object information, a configuration category to be issued, a configuration parameter, and a configuration condition.

[0274] The above describes the knowledge unit information corresponding to different control loop phases.

[0275] The above describes the first knowledge. The following describes the first request message.

[0276] In an implementation manner, the first request message can be used to request structured knowledge.

[0277] For example, when the first request message requests structured knowledge, the first request message can include at least one of the following: knowledge type, knowledge range, knowledge information, and knowledge filtering information of the first knowledge.

[0278] The knowledge filtering information is used to represent knowledge that does not need to be reported. The knowledge type, knowledge range, and knowledge information of the first knowledge are described above and will not be repeated here.

[0279] In another implementation manner, the first request message can be used to request unstructured knowledge.

[0280] For example, the first request message can be a natural language description of a live network problem.

[0281] In S502, the first node receives the first response information.

[0282] The first response information includes the first knowledge.

[0283] In some embodiments, the first node creates the first instance based on the first response information. The first node sends a creation response information to a third node. The third node is a management service consumer node.

[0284] In some embodiments, the first node receives a creation request from the third node.

[0285] The third node is a management service consumer node, and the creation request is used to instruct the first node (i.e., a management service provider node) to create the first instance.

[0286] In some embodiments, the knowledge type corresponding to the first knowledge is an artificial intelligence / machine learning management type, and the first instance is an artificial intelligence / machine learning instance.

[0287] In some embodiments, the knowledge type corresponding to the first knowledge is a control closed loop type, and the first instance is a control closed loop instance.

[0288] In some embodiments, the first node determines the creation request based on the first response information. The first node sends the creation request to a fourth node. The first node receives a creation response information from the fourth node.

[0289] The creation request is used to create the first instance. The creation response information is determined by the fourth node after creating the first instance based on the creation request.

[0290] In some embodiments, the fourth node can be a management service provider node.

[0291] In an example, the process in which the first node determines the creation request based on the first response information can include: the first node modifies a parameter in the creation request based on the first response information to determine the creation request.

[0292] Based on the above technical solution, the present disclosure provides a knowledge management method. The first node sends first request information, and the first request information is used to request first knowledge. The first node receives first response information, and the first response information includes the first knowledge. Compared with the current situation that there is no scheme related to the utilization of management service knowledge, the above technical solution can effectively utilize knowledge and avoid resource waste.

[0293] As an embodiment of the present disclosure, FIG. 6 is a flowchart of another knowledge management method according to an embodiment of the present disclosure, which can be applied to a second node. As shown in FIG. 6, the knowledge management method includes the following S601 and S602.

[0294] In S601, the second node receives first request information.

[0295] The second node can be a knowledge management service producer node. The first request information is used to request the first knowledge.

[0296] For example, the first request information and the introduction of the first knowledge refer to the embodiment shown as S501, which will not be repeated here.

[0297] In S602, the second node sends the first response information.

[0298] The first response information includes the first knowledge.

[0299] In an implementation mode, as shown in FIG. 7, corresponding to the second node being a management service supply producer node, before S602, the second node performs a knowledge query to determine the first knowledge.

[0300] Based on the above technical solution, the present disclosure provides a knowledge management method, the second node receives the first request information. The second node sends the first response information. Compared with the current scheme without using the knowledge related to the management service, thereby causing the problem of resource waste, in the above technical solution, the second node can respond to the first request information and send the first response information, effectively utilize the knowledge in the second node, and avoid resource waste.

[0301] As an embodiment of the present disclosure, FIG. 7 is a flow chart of another knowledge management method according to an embodiment of the present disclosure, which can be applied to a first node. As shown in FIG. 7, the knowledge management method includes the following S701 and S702.

[0302] In S701, the first node sends second request information.

[0303] The second request information is used to request to perform an archiving operation on the second knowledge. The second request information is also used to request to perform an updating operation on the second knowledge.

[0304] In some embodiments, the second knowledge includes structured knowledge and / or unstructured knowledge. Corresponding to the structured knowledge, the second knowledge includes at least one of the following: knowledge type, knowledge range, and knowledge information. Corresponding to the unstructured knowledge, the second knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0305] For example, the second knowledge can refer to the related introduction of the first knowledge in S501, which will not be repeated here.

[0306] Further, after receiving the second request message, the second node archives the second knowledge according to the second request message, corresponding to the second knowledge in the received second request message being structured knowledge; and the second node archives the second knowledge according to an embedding model, corresponding to the second knowledge in the received second request message being unstructured knowledge.

[0307] In S702, the first node receives second response information.

[0308] The second response information includes an archive state of the second knowledge.

[0309] For example, the archive state of the second knowledge can be knowledge archive completion, knowledge archive in progress, or knowledge archive failure.

[0310] Based on the above technical solution, the first node sends the second request information, and the first node receives the second response information. Compared with the current situation that there is no solution for utilizing knowledge related to management services, thereby causing the problem of resource waste, in the above technical solution, the first node sends the second request information to request the second node to perform an archive operation on the second knowledge, which facilitates subsequent utilization of the knowledge and improves the operation and management efficiency and avoids resource waste.

[0311] As an embodiment of the present disclosure, FIG. 8 is a flowchart of another knowledge management method according to an embodiment of the present disclosure, which can be applied to a second node. As shown in FIG. 8, the knowledge management method includes the following S801 and S802.

[0312] In S801, the second node receives second request information.

[0313] The second request information is used to request an archive operation on the second knowledge of the first instance. The second request information is also used to request an update operation on the second knowledge of the first instance.

[0314] For example, the second request information can refer to the embodiment shown in S701, which will not be described here.

[0315] In S802, the second node sends second response information.

[0316] The second response information includes an archive state of the second knowledge.

[0317] For example, the second response information can refer to the embodiment shown in S702, which will not be described here.

[0318] Based on the technical solution, the second node receives second request information, and the second node sends second response information. Compared with the current scheme without using knowledge related to the management service, the problem of resource waste is generated. In the technical solution, the second node can perform an archiving operation on the second knowledge in response to the second request information, facilitating subsequent use of the knowledge and avoiding resource waste.

[0319] As an embodiment of the present disclosure, FIG. 9 is a flowchart of another knowledge management method according to an embodiment of the present disclosure, which can be applied to a third node. The third node is a management service provider consumer. As shown in FIG. 9, the knowledge management method includes the following S901.

[0320] In S901, the third node sends a creation request to a fourth node.

[0321] The creation request is used to instruct the fourth node to create a control closed loop instance based on the first knowledge. The fourth node is a management service provider node. The control closed loop instance includes a control closed loop monitoring instance and / or a control closed loop management instance.

[0322] For example, the parameters of the control closed loop monitoring instance include at least one of a control closed loop monitoring indicator, a control closed loop monitoring strategy, and a control closed loop target.

[0323] The control closed loop monitoring indicator can represent various parameters, such as performance measurements, key performance indicators, customer experience indicators, key experience indicators, and key quality indicators. The control closed loop monitoring strategy is a monitoring strategy for different indicators, such as monitoring time, monitoring frequency, and reporting conditions. The control closed loop target is used to indicate the target threshold range of the indicators that need to be monitored and guaranteed in the control closed loop. After knowledge query, the indicators, indicator values, and monitoring strategies can be recommended and matched according to different guarantee scenarios.

[0324] For example, the parameters of the control closed loop management instance include at least one of a control closed loop monitoring indicator, a control closed loop monitoring strategy, a control closed loop target, a control closed loop priority, a control closed loop category, a control closed loop business category, a control closed loop component, a control closed loop object, and a control closed loop action condition. In some embodiments, after S901, the third node receives creation response information from the fourth node to determine whether the fourth node creates the control closed loop instance based on the creation request.

[0325] The creation response information includes parameters of the control closed loop instance, problems monitored for the control closed loop instance, and processing suggestions corresponding to the problems.

[0326] In some embodiments, the third node monitors parameters in the control loop monitoring instance and / or the control loop management instance, monitors control loop monitoring indicators, and creates a control loop report.

[0327] The control loop report includes at least one of a control loop operation report, a control loop failure report, and a control loop conflict report.

[0328] For example, the control loop failure report includes at least one of abnormal indicator information, failure category information, failure entity information, failure location information, failure cause information, and a failure recommended solution.

[0329] For example, the abnormal indicator information can be an abnormal key performance indicator; the failure category information can be a physical failure, a configuration error, a resource shortage, or a software failure; the failure entity information can be an identifier of a network function, an identifier of a network element, or an identifier of a multi-system network; the failure location information can be an access network, a core network, or an operation and maintenance management domain, or a device; the failure cause information can be an excessive number of terminals, traffic overload, or an excessive increase in the number of unlimited resource connections; and the failure recommended solution can be a failure handling scheme and an entity to which the scheme applies.

[0330] For example, the control loop conflict report includes at least one of a conflict category, a conflict cause, and a conflict recommended solution.

[0331] For example, the conflict category includes a control loop target, a control loop component, a control loop measure, and a control loop object. The conflict recommended solution includes a recommended replacement control loop component, an adjusted control loop target range, and a recommended control loop measure.

[0332] For example, the control loop operation report includes at least one of parameters of the control loop instance, control loop monitoring indicators, control loop target completion information, and failure handling information.

[0333] For example, the control loop monitoring indicators are numerical values of the parameters of the control loop instance, such as monitored indicator values. The control loop target completion information can be a completion percentage or information about whether the target is completed. The failure handling information can be a failure handling percentage or information about whether the failure handling is completed.

[0334] In some embodiments, the third node sends modification request information to a fourth node. The third node receives modification response information from the fourth node.

[0335] The modification request is used to indicate that the control loop instance is modified according to the control loop report. The modification response information can be a modification success response or a modification failure response. The modification failure response information includes a failure cause, such as a control loop component conflict.

[0336] In an example, the third node modifies the control loop target indicator range in the modification request information according to the control loop report and the first response information, and replaces the control loop component to determine the modification request information.

[0337] It can be understood that the third node sends the modification request information to the fourth node, and the third node receives the modification response information from the fourth node. The above technical solution can realize passive modification of the control loop instance, i.e., the third node instructs the fourth node to modify the control loop instance.

[0338] Based on the above technical solution, the third node sends a creation request to the fourth node. The creation request in the above technical solution can instruct the fourth node to create the control loop instance based on the first knowledge.

[0339] As an embodiment of the present disclosure, FIG. 10 is a flowchart of another knowledge management method according to an embodiment of the present disclosure, which can be applied to the fourth node. The fourth node is a management service provider. As shown in FIG. 10, the knowledge management method includes the following S1001.

[0340] In S1001, the fourth node receives a creation request.

[0341] The creation request is used to instruct the fourth node to create the control loop instance based on the first knowledge. The fourth node is a management service provider node. The control loop instance includes a control loop monitoring instance and / or a control loop management instance.

[0342] For example, the parameters of the control loop monitoring instance include at least one of a control loop monitoring indicator, a control loop monitoring strategy, and a control loop target.

[0343] For example, the parameters of the control loop management instance include at least one of a control loop monitoring indicator, a control loop monitoring strategy, a control loop target, a control loop priority, a control loop category, a control loop service category, a control loop component, a control loop object, and a control loop action condition.

[0344] For the introduction of each parameter, refer to the embodiment shown in S901, which will not be repeated here.

[0345] In some embodiments, the fourth node creates the control loop instance based on the creation request. The fourth node sends a creation response information to notify the third node whether the control loop instance has been created based on the creation request.

[0346] The creation response information includes the parameters of the control loop instance, the problem to be monitored for the control loop instance, and the processing suggestion corresponding to the problem.

[0347] In some embodiments, the fourth node receives modification request information. The fourth node modifies the control loop instance according to the control loop report based on the modification request information. The fourth node sends modification response information.

[0348] The modification request information is used to indicate modification of the control loop instance according to the control loop report.

[0349] For example, the modification request information and the modification response information refer to the embodiment shown in S901, which will not be described here.

[0350] It can be understood that the fourth node receives modification request information; the fourth node modifies the control loop instance according to the control loop report based on the modification request information; and the fourth node sends modification response information. The above technical solution can realize passive modification of the control loop instance, i.e., the third node instructs the fourth node to modify the control loop instance.

[0351] Based on the above technical solution, the fourth node receives a creation request. In the above technical solution, the fourth node can receive the creation request so as to create the control loop instance based on the first knowledge.

[0352] As an embodiment of the present disclosure, FIG. 11 is a flowchart of another knowledge management method according to an embodiment of the present disclosure, which can be applied to the fourth node. The fourth node is a management service provider. As shown in FIG. 11, the knowledge management method includes the following S1101 and S1102.

[0353] In S1101, the fourth node creates a control loop instance based on the first knowledge.

[0354] In an example, the fourth node creates a first instance of a corresponding type according to the knowledge type of the first knowledge. For example, in the case where the knowledge type of the first knowledge is a control loop type, a control loop instance is created.

[0355] In another example, in the case where a dynamic control loop creation request of the third node is received, the fourth node creates a dynamic control loop instance based on the first knowledge and the dynamic control loop creation request.

[0356] In S1102, the fourth node sends creation response information to the third node.

[0357] The third node is a management service consumer node. The creation response information includes parameters of the control loop instance, a problem monitored for the control loop instance, and a processing suggestion corresponding to the problem.

[0358] In some embodiments, the fourth node monitors a control loop monitoring index according to parameters in the control loop monitoring instance and / or the control loop management instance, and creates a control loop report. The fourth node sends the first request information. The fourth node receives the first response information.

[0359] Further, the fourth node sends the first request information. The fourth node receives the first response information.

[0360] For example, the first request information can refer to the embodiment shown in S501, which will not be repeated here. The first response information can refer to the embodiment shown in S502, which will not be repeated here.

[0361] In some embodiments, the fourth node receives a dynamic control closed loop creation request from the third node.

[0362] The dynamic control closed loop request includes one of the following parameters: dynamic control closed loop indication information, used to indicate that the control closed loop is a dynamic control closed loop; dynamic control closed loop component information; dynamic control closed loop monitoring index; and dynamic control closed loop target.

[0363] It can be understood that the dynamic control closed loop creation request is used to request to create a dynamic control closed loop for game service guarantee of a certain area. The dynamic control closed loop can dynamically adjust the content in the control closed loop instance, such as the control closed loop component and the control closed loop target.

[0364] In some embodiments, the fourth node modifies the control closed loop instance according to the monitoring of the control closed loop monitoring index. The fourth node sends a dynamic control closed loop report to the third node.

[0365] The dynamic control closed loop report contains the modified control closed loop component and the control closed loop target.

[0366] In an example, the fourth node dynamically modifies the control closed loop target index range in the control closed loop instance and / or replaces the control closed loop component according to the monitoring of the control closed loop monitoring index.

[0367] Further, the fourth node reselects and configures the control closed loop component according to the modified control closed loop instance.

[0368] Based on the above technical solution, the fourth node creates a control closed loop instance based on the first knowledge, and sends a creation response information to the third node. The above technical solution can create a control closed loop instance based on the first knowledge, which realizes effective utilization of the first knowledge and avoids resource waste.

[0369] As an embodiment of the present disclosure, FIG. 12 is a flowchart of another knowledge management method according to an embodiment of the present disclosure, which can be applied to a third node. The third node is a management service provider consumer node. As shown in FIG. 12, the knowledge management method includes the following S1201.

[0370] In S1201, the third node receives a creation response information.

[0371] For example, the creation response information is created according to the description of the creation response information in the embodiment shown in FIG. 18, which is not repeated here.

[0372] In some embodiments, the third node sends a dynamic control loop creation request to the fourth node.

[0373] For example, the dynamic control loop creation request is according to the embodiment shown in S1102, which is not repeated here.

[0374] In some embodiments, the third node receives a dynamic control loop report.

[0375] For example, the dynamic control loop report is according to the embodiment shown in S1102, which is not repeated here.

[0376] It can be understood that the third node receives the dynamic control loop report to determine the information of the dynamically modified control loop component and the control loop target in time.

[0377] Based on the above technical solution, the third node receives the creation response information to determine whether the control loop instance has been successfully created in time.

[0378] In some embodiments, as shown in FIG. 13, the management service provider node receives a creation request of a management service instance from a management service consumer node. The management service provider node sends first request information to a knowledge management service provider node. Correspondingly, the knowledge management service provider node receives the first request information. The knowledge management service provider node performs knowledge query. The knowledge management service provider node sends first response information. Correspondingly, the management service provider node receives the first response information.

[0379] Further, the management service provider node creates a management service instance according to the first response information. The management service provider node sends creation response information to the management service consumer node.

[0380] The management service provider node is the first node, the management service consumer node is the third node, the knowledge management service provider node is the second node, and the management service instance is the first instance.

[0381] In some embodiments, as shown in FIG. 14, the management service consumer node sends first request information to the knowledge management service provider node. Correspondingly, the knowledge management service provider node receives the first request information. The knowledge management service provider node performs knowledge query. The knowledge management service provider node sends first response information. Correspondingly, the management service consumer node receives the first response information.

[0382] Further, the management service provision consumer node modifies parameters in the management service provision creation request according to the first response information, and determines the management service provision creation request. The management service provision consumer node sends the management service provision creation request to the management service provision producer node. Correspondingly, the management service provision producer node receives the management service provision creation request. The management service provision producer node creates the management service provision instance. The management service provision producer node sends the creation response information to the management service provision consumer node.

[0383] The management service provision consumer node is the first node, the management service provision producer node is the fourth node, the knowledge management service producer node is the second node, and the management service provision instance is the first instance.

[0384] In some embodiments, as shown in FIG. 15, the management service provision consumer node sends the management service provision creation request to the management service provision producer node. Correspondingly, the management service provision producer node receives the management service provision creation request. The management service provision producer node creates the management service provision instance.

[0385] Further, the management service provision producer node sends the second request information to the knowledge management service producer node. Correspondingly, the knowledge management service producer node receives the second request information. The knowledge management service producer node performs the knowledge archiving. The knowledge management service producer node sends the second response information. Correspondingly, the management service provision producer node receives the second response information. The management service provision producer node sends the creation response information to the management service provision consumer node.

[0386] The management service provision producer node is the first node, the management service provision consumer node is the third node, the knowledge management service producer node is the second node, and the management service provision instance is the first instance.

[0387] In some embodiments, as shown in FIG. 16, the management service provision consumer node sends the management service provision creation request to the management service provision producer node. Correspondingly, the management service provision producer node receives the management service provision creation request. The management service provision producer node creates the management service provision instance. The management service provision producer node sends the creation response information to the management service provision consumer node.

[0388] Further, the management service provision consumer node sends the second request information to the knowledge management service producer node. Correspondingly, the knowledge management service producer node receives the second request information. The knowledge management service producer node performs the knowledge archiving. The knowledge management service producer node sends the second response information. Correspondingly, the management service provision consumer node receives the second response information.

[0389] The management service supply consumer node is the first node, the management service supply producer node is the fourth node, the knowledge management service producer node is the second node, and the management service supply instance is the first instance.

[0390] In an implementation, the first node creates a first instance of a corresponding type according to the knowledge type of the first knowledge in the first response information. For example, in a case where the knowledge type of the first knowledge is an artificial intelligence / machine learning management type, an artificial intelligence / machine learning instance is created; in a case where the knowledge type of the first knowledge is a control loop type, a control loop instance is created.

[0391] In some embodiments, the knowledge type corresponding to the first knowledge is an artificial intelligence / machine learning management type, and the first instance is an artificial intelligence / machine learning instance. The first node (the management service supply producer node) is a machine learning producer node.

[0392] In some embodiments, the knowledge type corresponding to the first knowledge is a control loop type, and the first instance is a control loop instance. The first node (the management service supply producer node) is a control loop producer node.

[0393] In some embodiments, as shown in FIG. 17, the machine learning producer node receives a machine learning training request (i.e., a creation request) from the machine learning consumer node. The machine learning producer node sends first request information to the knowledge management service producer node. Accordingly, the knowledge management service producer node receives the first request information. The knowledge management service producer node sends first response information. Accordingly, the machine learning producer node receives the first response information. The machine learning producer node collects and processes training data according to the first response information, and performs machine learning training to create a first instance. For example, re-collecting training data, pruning data to adjust the statistical distribution of data, adjusting machine learning training strategies, etc.

[0394] Further, the machine learning producer node sends creation response information to the machine learning consumer node, including the situation of machine learning training. The machine learning producer node archives knowledge of the training data (i.e., second knowledge), including: the machine learning producer node sends second request information to the knowledge management service producer node. The knowledge management service producer node performs knowledge archiving. The knowledge management service producer node sends second response information to the machine learning producer node.

[0395] The machine learning consumer node is the third node. The machine learning producer node is the first node. The machine learning training request includes an artificial intelligence / machine learning inference name (e.g., management data analysis type, analysis identifier, etc. information). The steps of knowledge archiving refer to the embodiments shown in S701 and S702, which will not be described here.

[0396] For example, the knowledge type of the first knowledge in the first request message is used to indicate the type of the first knowledge of the query, and in this embodiment, the knowledge type of the first knowledge in the first request message is an artificial intelligence / machine learning management type. Correspondingly, in the case where the knowledge type of the first knowledge is an artificial intelligence / machine learning management type, the knowledge information is referred to the above introduction of the first knowledge, and will not be repeated here. The knowledge scope and the knowledge filtering information are referred to the above introduction of the first request information, and will not be repeated here.

[0397] It should be noted that the first knowledge requested by the first request message can not exist in a certain machine learning workflow stage, that is, the first request message can not indicate a machine learning workflow stage when used to query the first knowledge, and can request machine learning knowledge information. Of course, the first request message can also be used to request knowledge unit information of the machine learning workflow stage.

[0398] In some embodiments, as shown in FIG. 18, the control loop consumer node sends a control loop creation request (i.e., a creation request). Correspondingly, the control loop producer node receives the control loop creation request from the control loop consumer node. The control loop producer node sends the first request information to the knowledge management service producer node. Correspondingly, the knowledge management service producer node receives the first request information. The knowledge management service producer node sends the first response information.

[0399] The control loop creation request is used to request to create a control loop for a certain regional game service guarantee.

[0400] At this time, the third node is a control loop consumer node. The first node is a control loop producer node, and the management service capability is a control loop or a communication service guarantee.

[0401] The control loop creation request (i.e., the creation request) includes at least one of a control loop priority, a control loop category, a control loop service category, a control loop component, a control loop object, a control loop target, and a control loop action condition.

[0402] The control loop priority indicates the priority of the control loop, which can be used to handle control loop conflicts.

[0403] The control loop category is used to indicate the category of the service on which the control loop acts, including service guarantee, fault handling, complaint handling, and / or service quality optimization.

[0404] The control loop service category includes at least one of voice, data (game, webpage, video, video and data service business), and message.

[0405] The control closed loop component information includes at least one of a terminal, a network element (for example, a multi-system network facility, a management service, a network function, a cell, an access network device), a transport network element, an application, and a service provider.

[0406] The control closed loop object includes at least one of a network slice, a terminal (or multiple terminals), a cell (or multiple cells), and a service.

[0407] The control closed loop target indicates the indicators that need to be monitored and guaranteed by the control closed loop and the threshold range of the guarantee, including but not limited to performance measurement, customer experience indicator, key experience indicator, key performance indicator, and key quality indicator.

[0408] The control closed loop action condition includes but is not limited to time condition (for example, specific time), area condition (for example, specific area), and threshold condition (for example, threshold of certain key performance indicator).

[0409] In some embodiments, the first request information can include the knowledge type (for example, control closed loop type), knowledge range (cross-domain knowledge, service guarantee knowledge), and knowledge information (for example, at least one of control closed loop category (for example, service guarantee), control closed loop service category (for example, data-game), control closed loop component information (for example, multi-system network facility, management service, network function, cell, and access network device), and control closed loop corresponding knowledge unit information) of the first knowledge.

[0410] It can be understood that the first request information can query the corresponding knowledge unit information for a control closed loop, or can query the corresponding knowledge unit information for a certain control closed loop stage, which is not limited in the present disclosure.

[0411] In the case of querying the corresponding knowledge unit information for a certain control closed loop stage, the knowledge unit information corresponding to different control closed loop stages requested in the first request information refers to the embodiments shown in Examples 21 to 23 described above, which will not be described here.

[0412] Further, after the knowledge management service producer node sends the first response information, the control closed loop producer node creates a control closed loop monitoring instance and / or a control closed loop management instance through the control closed loop component. The control closed loop producer node sends a creation response information to the control closed loop consumer node.

[0413] The creation response information includes parameters of the control closed loop instance, problems monitored for the control closed loop instance, and processing suggestions corresponding to the problems.

[0414] For example, the parameters of the control closed loop monitoring instance include at least one of control closed loop monitoring indicators, control closed loop monitoring strategies, and control closed loop targets.

[0415] The control closed loop monitoring indicator can embody the values of various parameters. The control closed loop target is used to indicate the indicator machine threshold range that needs to be monitored and guaranteed in the control closed loop, such as a customer experience indicator, a key experience indicator, a key performance indicator, and a key quality indicator. After knowledge query, the indicators can be recommended and matched according to different guarantee scenarios.

[0416] For example, the parameters of the control closed loop management instance include at least one of a control closed loop monitoring indicator, a control closed loop monitoring strategy, a control closed loop target, a control closed loop priority, a control closed loop category, a control closed loop business category, a control closed loop component, a control closed loop object, and a control closed loop action condition.

[0417] In some embodiments, as shown in FIG. 19, the control closed loop consumer node sends a control closed loop creation request to the control closed loop producer node. Accordingly, the control closed loop producer node receives the control closed loop creation request of the control closed loop consumer node. The control closed loop producer node creates a control closed loop instance through a control closed loop component. The control closed loop producer node sends creation response information to the control closed loop consumer node. Accordingly, the control closed loop consumer node receives the creation response information.

[0418] Further, the control closed loop producer node monitors the parameters of the control closed loop monitoring instance through the control closed loop component, detects whether the control closed loop monitoring instance has a control closed loop failure or a control closed loop conflict, and sends a control closed loop failure report and / or a control closed loop conflict report to the control closed loop consumer node.

[0419] Further, the control closed loop consumer node sends first request information to the knowledge management service producer node and receives first response information. The control closed loop consumer node sends modification request information to the control closed loop producer node. The control closed loop producer node modifies the control closed loop instance according to the control closed loop failure report and / or the control closed loop conflict report. The control closed loop consumer node receives modification response information from the control closed loop producer node.

[0420] In some embodiments, as shown in FIG. 20, the control closed loop consumer node sends a dynamic control closed loop creation request to the control closed loop producer node. Accordingly, the control closed loop producer node receives the dynamic control closed loop creation request of the control closed loop consumer node. The control closed loop producer node creates a dynamic control closed loop through a control closed loop component. The control closed loop producer node sends dynamic control closed loop creation response information to the control closed loop consumer node. Accordingly, the control closed loop consumer node receives the dynamic control closed loop creation response information.

[0421] The third node is a control closed loop consumer node, the second node is a knowledge management service producer node, and the fourth node is a control closed loop producer node.

[0422] Further, the control loop producer node monitors parameters of the control loop monitoring instance through the control loop component, detects whether the control loop monitoring instance has a control loop fault or a control loop conflict.

[0423] Further, the control loop producer node sends first request information to the knowledge management service producer node, and receives first response information. The control loop producer node modifies the dynamic control loop according to the knowledge of the control loop fault or the control loop conflict (for example, dynamically modifies the control loop target indicator range in the control loop instance, and / or replaces the control loop component).

[0424] Further, the control loop producer node performs control loop component reselection and configuration according to the modified control loop instance, and sends a dynamic control loop report to the control loop consumer node to report the modified control loop target indicator range and / or the replaced control loop component.

[0425] The embodiments of the present disclosure can divide the functional modules or functional units of the knowledge management apparatus according to the above-mentioned method examples. For example, each functional module or functional unit can be divided according to each function, or two or more functions can be integrated into one processing module. The integrated module can be implemented in the form of hardware or software functional modules or functional units. The division of modules or units in the embodiments of the present disclosure is illustrative, and is only a logical function division. When actually implemented, another division mode can be used.

[0426] As shown in FIG. 21, it is a structural schematic diagram of a knowledge management apparatus 210 according to an embodiment of the present disclosure, which comprises a communication unit 2101 and a processing unit 2102.

[0427] The communication unit 2101 is configured to send first request information; the first request information is used to request first knowledge.

[0428] The communication unit 2101 is further configured to receive first response information; the first response information comprises the first knowledge.

[0429] In an implementation mode, the first knowledge comprises structured knowledge and / or unstructured knowledge.

[0430] In an implementation mode, corresponding to the structured knowledge, the first knowledge comprises knowledge type, knowledge range, and / or knowledge information.

[0431] In an implementation mode, corresponding to the unstructured knowledge, the first knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0432] In an implementation, the knowledge type includes: an artificial intelligence / machine learning management type, and / or, a control loop type.

[0433] In an implementation, the knowledge type corresponding to the first knowledge is the artificial intelligence / machine learning management type, and the knowledge information includes: knowledge unit information of at least one machine learning workflow stage. The machine learning workflow stage includes at least one of: a machine learning model training stage, a machine learning model testing stage, an artificial intelligence / machine learning inference simulation stage, a machine learning model deployment stage, and an artificial intelligence / machine learning inference stage.

[0434] In an implementation, the machine learning workflow stage corresponding to the first knowledge is the machine learning model training stage, and the knowledge unit information includes: at least one of: training data information, training strategy information, and machine learning model information. The training data information includes at least one of: training data feature information, training data distribution information, training data volume information, and training data collection strategy. The training strategy information includes training energy consumption information, and / or training orchestration information. The machine learning model information includes at least one of: model energy consumption information, model complexity information, model size information, model performance information, and model confidence information.

[0435] In an implementation, the machine learning workflow stage corresponding to the first knowledge is the machine learning model testing stage, and the knowledge unit information includes: at least one of: test set to training set ratio information, testing data information, and testing strategy information.

[0436] In an implementation, the machine learning workflow stage corresponding to the first knowledge is the artificial intelligence / machine learning inference stage, and the knowledge unit information includes: at least one of: inference data information, inference strategy information, and machine learning model information. The inference data information includes at least one of: inference data feature information, inference data distribution information, inference data volume information, inference data collection strategy, and data collection condition. The inference strategy information includes inference energy consumption information, and / or inference orchestration information.

[0437] In an implementation, the knowledge type corresponding to the first knowledge is the control loop type, and the knowledge information includes: at least one of: control loop category, control loop business category, control loop component information, and control loop corresponding knowledge unit information.

[0438] In an implementation, the control loop corresponding knowledge unit information includes: knowledge unit information corresponding to at least one control loop stage. The control loop stage includes at least one of: monitoring stage, analysis stage, decision stage, and execution stage.

[0439] In an implementation, the control loop stage corresponding to the monitoring stage comprises at least one of a monitoring index and a monitoring strategy.

[0440] In an implementation, the control loop stage corresponding to the analysis stage or the decision stage comprises at least one of control loop fault information, control loop conflict information, and fault handling suggestions.

[0441] In an implementation, the control loop stage corresponding to the execution stage comprises scenario information, an execution scheme, an execution board, and execution flow information.

[0442] In an implementation, the first request message comprises at least one of a knowledge type of the first knowledge, a knowledge range, knowledge information, and knowledge filtering information.

[0443] In an implementation, the first node is a management service supply consumer node or a management service supply producer node.

[0444] In an implementation, the first node is the management service supply producer node, and the knowledge management method further comprises: creating the first instance based on the first response information; and sending a creation response message to a third node, the third node being the management service supply consumer node.

[0445] In an implementation, before sending the first request information, the communication unit 2401 is further configured to receive a creation request from a third node, the creation request being used for the management service supply producer node to create the first instance.

[0446] In an implementation, the first node is the management service supply consumer node, and the processing unit 2102 is configured to determine a creation request based on the first response information, the creation request being used for creating the first instance.

[0447] The communication unit 2101 is further configured to send a creation request to a fourth node, the fourth node being the management service supply producer node, and receive a creation response message from the fourth node.

[0448] In an implementation, the knowledge type of the first knowledge is an artificial intelligence / machine learning management type, and the first instance is an artificial intelligence / machine learning instance.

[0449] In an implementation, the knowledge type of the first knowledge is a control loop type, and the first instance is a control loop instance.

[0450] In an implementation manner, the communication unit 2101 is further configured to send second request information, the second request information being used to request to perform an archiving operation on the second knowledge of the first instance; and receive second response information, the second response information including an archiving state of the second knowledge.

[0451] In an implementation manner, the second knowledge includes structured knowledge, and / or unstructured knowledge.

[0452] In an implementation manner, corresponding to the structured knowledge, the second knowledge includes at least one of the following: a knowledge type, a knowledge range, and knowledge information.

[0453] In an implementation manner, corresponding to the unstructured knowledge, the second knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0454] In an implementation manner, the knowledge management apparatus 210 can further include a storage unit 2103 (shown in a dashed box in FIG. 21) that stores programs or instructions. When the processing unit 2102 executes the programs or instructions, the knowledge management apparatus 210 can perform the knowledge management method described in the above method embodiments.

[0455] As shown in FIG. 22, it is a structural schematic diagram of another knowledge management apparatus 220 according to an embodiment of the present disclosure. The knowledge management apparatus 220 includes a communication unit 2201.

[0456] The communication unit 2201 is configured to receive first request information, the first request information being used to request first knowledge.

[0457] The communication unit 2201 is further configured to send first response information, the first response information including the first knowledge.

[0458] In an implementation manner, the first knowledge includes structured knowledge, and / or unstructured knowledge.

[0459] In an implementation manner, corresponding to the structured knowledge, the first knowledge includes a knowledge type, a knowledge range, and / or knowledge information.

[0460] In an implementation manner, corresponding to the unstructured knowledge, the first knowledge is a knowledge vector and index information corresponding to the knowledge vector.

[0461] In an implementation manner, the knowledge type includes an artificial intelligence / machine learning management type, and / or a control closed loop type.

[0462] In an implementation, the knowledge type corresponding to the first knowledge is an artificial intelligence / machine learning management type, and the knowledge information includes knowledge unit information of at least one machine learning workflow stage. The machine learning workflow stage includes at least one of a machine learning model training stage, a machine learning model testing stage, an artificial intelligence / machine learning inference simulation stage, a machine learning model deployment stage, and an artificial intelligence / machine learning inference stage.

[0463] In an implementation, the machine learning workflow stage corresponding to the machine learning model training stage, the knowledge unit information includes at least one of training data information, training strategy information, and machine learning model information. The training data information includes at least one of training data feature information, training data distribution information, training data volume information, and training data collection strategy. The training strategy information includes training energy consumption information and / or training orchestration information. The machine learning model information includes at least one of model energy consumption information, model complexity information, model size information, model performance information, and model confidence information.

[0464] In an implementation, the machine learning workflow stage corresponding to the machine learning model testing stage, the knowledge unit information includes at least one of test set to training set ratio information, test data information, and test strategy information.

[0465] In an implementation, the machine learning workflow stage corresponding to the artificial intelligence / machine learning inference stage, the knowledge unit information includes at least one of inference data information, inference strategy information, and machine learning model information. The inference data information includes at least one of inference data feature information, inference data distribution information, inference data volume information, inference data collection strategy, and data collection condition. The inference strategy information includes inference energy consumption information and / or inference orchestration information.

[0466] In an implementation, the knowledge type corresponding to the first knowledge is a control closed loop type, and the knowledge information includes at least one of control closed loop category, control closed loop business category, control closed loop component information, and control closed loop corresponding knowledge unit information.

[0467] In an implementation, the control closed loop corresponding knowledge unit information includes knowledge unit information corresponding to at least one control closed loop stage. The control closed loop stage includes at least one of a monitoring stage, an analysis stage, a decision stage, and an execution stage.

[0468] In an implementation, the control closed loop stage corresponding to the monitoring stage, the control closed loop stage corresponding knowledge unit information includes at least one of monitoring indicators and monitoring strategy.

[0469] In an implementation manner, the control closed loop stage corresponds to an analysis stage or a decision stage, and the knowledge unit information corresponding to the control closed loop stage comprises at least one of control closed loop fault information, control closed loop conflict information, and fault processing suggestion.

[0470] In an implementation manner, the control closed loop stage corresponds to an execution stage, and the knowledge unit information corresponding to the control closed loop stage comprises scene information, execution scheme, execution board, and execution flow information.

[0471] In an implementation manner, the first request message comprises at least one of knowledge type, knowledge range, knowledge information, and knowledge filtering information of the first knowledge.

[0472] In an implementation manner, the communication unit 2201 is further configured to receive second request information, and the second request information is used to request to perform an archiving operation on the second knowledge of the first instance; and send second response information, and the second response information comprises an archiving state of the second knowledge.

[0473] In an implementation manner, the second knowledge comprises structured knowledge and / or unstructured knowledge.

[0474] In an implementation manner, the second knowledge comprises at least one of knowledge type, knowledge range, and knowledge information corresponding to the structured knowledge.

[0475] In an implementation manner, the second knowledge is a knowledge vector and index information corresponding to the knowledge vector corresponding to the unstructured knowledge.

[0476] In an implementation manner, the knowledge management apparatus 220 can further comprise a storage unit 2202 (shown in a dashed box in FIG. 22) storing programs or instructions. When the communication unit 2201 executes the programs or instructions, the knowledge management apparatus 220 can execute the knowledge management method described in the above method embodiments.

[0477] As shown in FIG. 23, it is a structural schematic diagram of another knowledge management apparatus 230 according to an embodiment of the present disclosure. The knowledge management apparatus 230 comprises a communication unit 2301 and a processing unit 2302.

[0478] The communication unit 2301 is configured to send a creation request to a fourth node, and the creation request is used to instruct the fourth node to create a control closed loop instance based on first knowledge. The fourth node is a management service supply producer node.

[0479] In an implementation manner, the control closed loop instance comprises a control closed loop monitoring instance and / or a control closed loop management instance.

[0480] In an implementation manner, the parameter of the control loop monitoring instance comprises at least one of a control loop monitoring index, a control loop monitoring strategy, and a control loop target.

[0481] In an implementation manner, the parameter of the control loop management instance comprises at least one of a control loop monitoring index, a control loop monitoring strategy, a control loop target, a control loop priority, a control loop category, a control loop service category, a control loop component, a control loop object, and a control loop action condition.

[0482] In an implementation manner, the communication unit 2301 is further configured to receive, from the fourth node, creation response information. The creation response information comprises at least one of the parameter of the control loop instance, a problem to be monitored for the control loop instance, and a processing suggestion corresponding to the problem.

[0483] In an implementation manner, the processing unit 2302 is configured to monitor a control loop monitoring index according to the parameter in the control loop monitoring instance and / or the control loop management instance, and create a control loop report. The control loop report comprises at least one of a control loop operation report, a control loop fault report, and a control loop conflict report.

[0484] In an implementation manner, the control loop fault report comprises at least one of abnormal index information, fault category information, fault entity information, fault positioning information, fault reason information, and fault recommended solution.

[0485] In an implementation manner, the control loop conflict report comprises at least one of a conflict category, a conflict reason, and a conflict recommended solution.

[0486] In an implementation manner, the control loop operation report comprises at least one of each parameter of the control loop instance, a control loop monitoring index, a control loop target completion condition, and a fault processing condition.

[0487] In an implementation manner, the communication unit 2301 is further configured to send, to the fourth node, modification request information, the modification request being used to indicate that the control loop instance is modified according to the control loop report; and receive, from the fourth node, modification response information.

[0488] In an implementation manner, the knowledge management apparatus 230 can further comprise a storage unit 2303 (shown in a dashed box in FIG. 23) storing programs or instructions. When the processing unit 2302 executes the programs or instructions, the knowledge management apparatus 230 can perform the knowledge management method described in the above method embodiments.

[0489] As shown in FIG. 24, it is a structural schematic diagram of still another knowledge management apparatus 240 according to an embodiment of the present disclosure. The knowledge management apparatus 240 comprises a communication unit 2401 and a processing unit 2402.

[0490] The processing unit 2402 is configured to create a control loop instance based on the first knowledge. The communication unit 2401 is configured to send a creation response information to a third node. The third node is a management service provider node.

[0491] In an implementation, the communication unit 2401 is further configured to receive a dynamic control loop creation request from the third node. The dynamic control loop creation request comprises one of the following parameters: dynamic control loop indication information, used to indicate that the control loop is a dynamic control loop; dynamic control loop component information; dynamic control loop monitoring index; and dynamic control loop target.

[0492] In an implementation, the processing unit 2402 is further configured to modify the control loop instance according to a monitoring result of the control loop monitoring index. The communication unit 2401 is further configured to send a dynamic control loop report to the third node.

[0493] In an implementation, the knowledge management apparatus 240 can further comprise a storage unit 2403 (shown in a dashed box in FIG. 24), which stores programs or instructions. When the processing unit 2402 executes the programs or instructions, the knowledge management apparatus 240 can perform the knowledge management method described in the above method embodiments.

[0494] From the above description of the implementations, those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional modules is taken as an example. In actual applications, the above functions can be completed by different functional modules according to needs, i.e., the internal structure of the apparatus is divided into different functional modules to complete all or part of the functions described above. The working processes of the above-described system, apparatus and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0495] The embodiment of the present disclosure provides a computer program product containing instructions, which, when executed on a computer, cause the computer to perform the knowledge management method in the above method embodiments.

[0496] The embodiment of the present disclosure also provides a computer readable storage medium (for example, a non-transitory computer readable storage medium), which stores instructions, and when the instructions are executed on a computer, cause the computer to perform the knowledge management method in the method flow shown in the above method embodiments.

[0497] The computer readable storage medium, for example, can be, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Examples (a non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a register, a hard disk, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing, or any other medium from which a computer can read instructions. An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. Of course, the storage medium can be a part of the processor. The processor and the storage medium can be located in an application specific integrated circuit (ASIC). In the embodiments of the present disclosure, the computer readable storage medium can be any tangible medium that contains or stores a program used by or in connection with an instruction execution system, apparatus, or device.

[0498] Since the knowledge management apparatus, the computer readable storage medium, and the computer program product in the embodiments of the present disclosure can be applied to the above method, the technical effects that can be achieved thereby can also be referred to the method embodiments, and the embodiments of the present disclosure will not be described here.

[0499] Through the technical solutions provided in the present disclosure, compared with the current situation that there is no solution for utilizing knowledge related to management services, thereby causing the problem of resource waste, the first node sends a request message and receives first response information including first knowledge, which can effectively utilize knowledge and avoid resource waste.

[0500] In the embodiments provided by the present disclosure, it should be understood that the disclosed system, device and method can be implemented in other manners. For example, the embodiments of the device described above are merely schematic, and the division of the units is merely logical function division, and there can be other division manners in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, or the among different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0501] It should be understood that the various forms of flow shown above can be reordered, added, or deleted steps. For example, the steps described in the present disclosure can be executed in parallel, in sequence, or in a different order, as long as the desired results of the technical solutions of the present disclosure can be achieved, and the present disclosure is not limited herein.

[0502] The specific implementation described above does not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent replacement and improvement made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

[0503] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they can be located in one place, or distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0504] In addition, each functional unit in each embodiment of the present disclosure can be integrated into a processing unit, or each unit can exist physically, or two or more units can be integrated into one unit.

[0505] The above is merely a specific implementation of the present disclosure, but the protection scope of the present disclosure is not limited thereto, and any change or replacement within the technical scope disclosed by the present disclosure shall be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. A knowledge management method applied to a first node, comprising: sending a first request information; the first request information is used to request a first knowledge; receiving a first response information; the first response information includes the first knowledge.

2. The method of claim 1, wherein, The first knowledge includes structured knowledge, and / or unstructured knowledge.

3. The method of claim 2, wherein, Corresponding to the structured knowledge, the first knowledge includes knowledge type, knowledge scope, and / or knowledge information.

4. The method of claim 2, wherein, Corresponding to the unstructured knowledge, the first knowledge is a knowledge vector and index information corresponding to the knowledge vector.

5. The method of claim 3, wherein, The knowledge type includes: artificial intelligence / machine learning management type, and / or control loop type.

6. The method of claim 5, wherein, Corresponding to the knowledge type of the first knowledge is the artificial intelligence / machine learning management type, the knowledge information includes at least one knowledge unit information; Wherein, the machine learning workflow stage includes at least one of the following: machine learning model training stage, machine learning model testing stage, artificial intelligence / machine learning inference simulation stage, machine learning model deployment stage, artificial intelligence / machine learning inference stage.

7. The method of claim 6, wherein, Corresponding to the machine learning workflow stage is the machine learning model training stage, the knowledge unit information includes at least one of: training data information, training strategy information, and machine learning model information; Wherein, the training data information includes at least one of training data feature information, training data distribution information, training data amount information, and training data collection strategy; The training strategy information includes training energy consumption information, and / or training arrangement information; The machine learning model information includes at least one of model energy consumption information, model complexity information, model size information, model performance information, and model confidence information.

8. The method of claim 6, wherein, Corresponding to the machine learning workflow stage is the machine learning model testing stage, the knowledge unit information includes at least one of: test set and training set ratio information, test data information, and test strategy information.

9. The method of claim 6, wherein, Corresponding to the machine learning workflow stage is the artificial intelligence / machine learning inference stage, the knowledge unit information includes at least one of: inference data information, inference strategy information, and machine learning model information; Wherein, the inference data information includes at least one of inference data feature information, inference data distribution information, inference data amount information, inference data collection strategy, and data collection condition; The inference strategy information includes inference energy consumption information, and / or inference arrangement information.

10. The method of claim 5, wherein, Corresponding to the knowledge type of the first knowledge is the control loop type, the knowledge information includes: control loop category, control loop business category, control loop component information, and control loop corresponding knowledge unit information.

11. The method of claim 10, wherein, The control loop corresponding knowledge unit information includes at least one control loop stage corresponding knowledge unit information; The control loop stage includes at least one of the following: monitoring stage, analysis stage, decision stage, and execution stage.

12. The method of claim 11, wherein, Corresponding to the control closed loop stage being the monitoring stage, the knowledge unit information corresponding to the control closed loop stage comprises at least one of a monitoring index and a monitoring strategy.

13. The method of claim 11, wherein, Corresponding to the control closed loop stage being the analysis stage or the decision stage, the knowledge unit information corresponding to the control closed loop stage comprises at least one of control closed loop fault information, control closed loop conflict information, and fault processing suggestions.

14. The method of claim 11, wherein, Corresponding to the control closed loop stage being the execution stage, the knowledge unit information corresponding to the control closed loop stage comprises scene information, an execution scheme, an execution board, and execution flow information.

15. The method of claim 1, wherein, The first request message comprises at least one of a knowledge type, a knowledge range, knowledge information, and knowledge filtering information of the first knowledge.

16. The method of claim 1, wherein, The first node is a management service supply consumer node or a management service supply producer node.

17. The method of claim 16, wherein, Corresponding to the first node being the management service supply producer node, the method further comprises: creating a first instance based on the first response information; sending a creation response information to a third node; the third node is a management service supply consumer node.

18. The method of claim 17, wherein, Before the sending of the first request information, the method further comprises: receiving a creation request from the third node; the creation request is used to instruct the management service supply producer node to create the first instance.

19. The method of claim 16, wherein, Corresponding to the first node being the management service supply consumer node, the method further comprises: determining a creation request based on the first response information; the creation request is used to create a first instance; sending the creation request to a fourth node; the fourth node is a management service supply producer node; receiving a creation response information from the fourth node.

20. The method of claim 17 or 19, wherein, Corresponding to the knowledge type of the first knowledge being an artificial intelligence / machine learning management type, the first instance is an artificial intelligence / machine learning instance.

21. The method of claim 17 or 19, wherein, Corresponding to the knowledge type of the first knowledge being a control closed loop type, the first instance is a control closed loop instance.

22. The method of claim 1, further comprising: sending a second request information; the second request information is used to request to perform an archiving operation on a second knowledge; receiving a second response information; the second response information comprises an archiving state of the second knowledge.

23. The method of claim 22, wherein, The second knowledge comprises structured knowledge, and / or unstructured knowledge.

24. The method of claim 23, wherein, Corresponding to the structured knowledge, the second knowledge comprises at least one of a knowledge type, a knowledge range, and knowledge information.

25. The method of claim 23, wherein, Corresponding to the unstructured knowledge, the second knowledge is a knowledge vector and index information corresponding to the knowledge vector.

26. A knowledge management method applied to a second node, comprising: receiving a first request information; the first request information is used to request a first knowledge; sending a first response information; the first response information comprises the first knowledge.

27. A knowledge management method applied to a third node, comprising: sending a creation request to a fourth node; the creation request is used to instruct the fourth node to create a control closed loop instance based on a first knowledge; the fourth node is a management service supply producer node.

28. The method of claim 27, wherein, The control loop instance comprises a control loop monitoring instance and / or a control loop management instance.

29. The method of claim 28, wherein, The parameters of the control loop monitoring instance comprise at least one of a control loop monitoring indicator, a control loop monitoring strategy, and a control loop target.

30. The method of claim 28, wherein, The parameters of the control loop management instance comprise at least one of a control loop monitoring indicator, a control loop monitoring strategy, a control loop target, a control loop priority, a control loop category, a control loop service category, a control loop component, a control loop object, and a control loop action condition.

31. The method of claim 27, further comprising: receiving a creation response information from the fourth node; the creation response information comprises parameters of the control loop instance, a problem to be monitored for the control loop instance, and a processing suggestion corresponding to the problem.

32. The method of claim 31, further comprising: monitoring a control loop monitoring indicator according to the parameters in the control loop monitoring instance and / or the control loop management instance, and creating a control loop report; the control loop report comprises at least one of a control loop operation report, a control loop fault report, and a control loop conflict report.

33. The method of claim 32, wherein, The control loop fault report comprises at least one of an abnormal indicator information, a fault category information, a fault entity information, a fault positioning information, a fault cause information, and a fault recommended solution.

34. The method of claim 32, wherein, The control loop conflict report comprises at least one of a conflict category, a conflict cause, and a conflict recommended solution.

35. The method of claim 32, wherein, The control loop operation report comprises at least one of parameters of the control loop instance, a control loop monitoring indicator, a control loop target completion, and a fault processing.

36. The method of claim 32, further comprising: sending a modification request information to the fourth node; the modification request information is used to indicate that the control loop instance is modified according to the control loop report; receiving a modification response information from the fourth node.

37. A knowledge management method applied to a fourth node, comprising: creating a control loop instance based on first knowledge; sending a creation response information to a third node; the third node is a management service supply consumer node.

38. The method of claim 37, further comprising: receiving a dynamic control loop creation request from the third node; the dynamic control loop request comprises one of the following parameters, dynamic control loop indication information, used to indicate that the control loop is a dynamic control loop; dynamic control loop component information; dynamic control loop monitoring indicator; dynamic control loop target.

39. The method of claim 38, further comprising: modifying the control loop instance according to the monitoring of the control loop monitoring indicator; sending a dynamic control loop report to the third node.

40. An electronic device, comprising: a processor and a communication interface; wherein the communication interface is coupled with the processor, and the processor is configured to run a computer program or instructions to implement the knowledge management method according to any one of claims 1-25, 26, 27-36, or 37-39.

41. A computer readable storage medium, wherein, The computer readable storage medium has stored therein instructions which, when executed by a computer, cause the computer to perform the knowledge management method according to any one of claims 1-25, claim 26, claims 27-36, or claims 37-39.

42. A computer program product, wherein, The computer program product comprises computer instructions which, when run on a computer, cause the computer to perform the knowledge management method according to any one of claims 1-25, claim 26, claims 27-36, or claims 37-39.

Images